Full Code of huseinzol05/Stock-Prediction-Models for AI

Repository: huseinzol05/Stock-Prediction-Models
Branch: master
Commit: 33266732b0b1
Files: 114
Total size: 14.9 MB

Directory structure:
gitextract_w7zipbjz/

├── .gitignore
├── LICENSE
├── README.md
├── agent/
│   ├── 1.turtle-agent.ipynb
│   ├── 10.duel-q-learning-agent.ipynb
│   ├── 11.double-duel-q-learning-agent.ipynb
│   ├── 12.duel-recurrent-q-learning-agent.ipynb
│   ├── 13.double-duel-recurrent-q-learning-agent.ipynb
│   ├── 14.actor-critic-agent.ipynb
│   ├── 15.actor-critic-duel-agent.ipynb
│   ├── 16.actor-critic-recurrent-agent.ipynb
│   ├── 17.actor-critic-duel-recurrent-agent.ipynb
│   ├── 18.curiosity-q-learning-agent.ipynb
│   ├── 19.recurrent-curiosity-q-learning-agent.ipynb
│   ├── 2.moving-average-agent.ipynb
│   ├── 20.duel-curiosity-q-learning-agent.ipynb
│   ├── 21.neuro-evolution-agent.ipynb
│   ├── 22.neuro-evolution-novelty-search-agent.ipynb
│   ├── 23.abcd-strategy-agent.ipynb
│   ├── 3.signal-rolling-agent.ipynb
│   ├── 4.policy-gradient-agent.ipynb
│   ├── 5.q-learning-agent.ipynb
│   ├── 6.evolution-strategy-agent.ipynb
│   ├── 7.double-q-learning-agent.ipynb
│   ├── 8.recurrent-q-learning-agent.ipynb
│   ├── 9.double-recurrent-q-learning-agent.ipynb
│   └── updated-NES-google.ipynb
├── dataset/
│   ├── AMD.csv
│   ├── BTC-sentiment.csv
│   ├── FB.csv
│   ├── FSV.csv
│   ├── GOOG-year.csv
│   ├── GOOG.csv
│   ├── INFY.csv
│   ├── KNX.csv
│   ├── MONDY.csv
│   ├── MTDR.csv
│   ├── SINA.csv
│   ├── TMUS.csv
│   ├── TSLA.csv
│   ├── TWTR.csv
│   ├── eur-myr.csv
│   ├── oil.csv
│   └── usd-myr.csv
├── deep-learning/
│   ├── 1.lstm.ipynb
│   ├── 10.lstm-seq2seq.ipynb
│   ├── 11.bidirectional-lstm-seq2seq.ipynb
│   ├── 12.lstm-seq2seq-vae.ipynb
│   ├── 13.gru-seq2seq.ipynb
│   ├── 14.bidirectional-gru-seq2seq.ipynb
│   ├── 15.gru-seq2seq-vae.ipynb
│   ├── 16.attention-is-all-you-need.ipynb
│   ├── 17.cnn-seq2seq.ipynb
│   ├── 18.dilated-cnn-seq2seq.ipynb
│   ├── 2.bidirectional-lstm.ipynb
│   ├── 3.lstm-2path.ipynb
│   ├── 4.gru.ipynb
│   ├── 5.bidirectional-gru.ipynb
│   ├── 6.gru-2path.ipynb
│   ├── 7.vanilla.ipynb
│   ├── 8.bidirectional-vanilla.ipynb
│   ├── 9.vanilla-2path.ipynb
│   ├── access.py
│   ├── addressing.py
│   ├── autoencoder.py
│   ├── dnc.py
│   ├── how-to-forecast.ipynb
│   ├── sentiment-consensus.ipynb
│   └── util.py
├── free-agent/
│   ├── evolution-strategy-agent.ipynb
│   └── evolution-strategy-bayesian-agent.ipynb
├── misc/
│   ├── bitcoin-analysis-lstm.ipynb
│   ├── fashion-forecasting.ipynb
│   ├── fashion.csv
│   ├── kijang-emas-bank-negara.ipynb
│   ├── outliers.ipynb
│   ├── overbought-oversold.ipynb
│   ├── sentiment-bitcoin.csv
│   ├── tesla-study.ipynb
│   └── which-stock.ipynb
├── realtime-agent/
│   ├── AMD.csv
│   ├── CPRT.csv
│   ├── FB.csv
│   ├── FSV.csv
│   ├── GOOG.csv
│   ├── GWR.csv
│   ├── LB.csv
│   ├── LYFT.csv
│   ├── MTDR.csv
│   ├── README.md
│   ├── SINA.csv
│   ├── TSLA.csv
│   ├── TWTR.csv
│   ├── app.py
│   ├── model.pkl
│   ├── realtime-evolution-strategy.ipynb
│   └── request.ipynb
├── simulation/
│   ├── monte-carlo-drift.ipynb
│   ├── monte-carlo-dynamic-volatility.ipynb
│   ├── monte-carlo-simple.ipynb
│   ├── multivariate-drift-monte-carlo.ipynb
│   └── portfolio-optimization.ipynb
├── stacking/
│   ├── autoencoder.py
│   ├── model.py
│   ├── stack-encoder-ensemble-xgb.ipynb
│   └── stack-rnn-arima-xgb.ipynb
└── stock-forecasting-js/
    ├── README.md
    ├── css/
    │   └── style.css
    ├── data/
    │   └── google.js
    ├── index.html
    ├── init.js
    └── js/
        ├── tf-expand.js
        ├── tf.js
        └── utils.js

================================================
FILE CONTENTS
================================================

================================================
FILE: .gitignore
================================================
*DS_Store
*ipynb_checkpoints
*__pycache__


================================================
FILE: LICENSE
================================================
                                 Apache License
                           Version 2.0, January 2004
                        http://www.apache.org/licenses/

   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION

   1. Definitions.

      "License" shall mean the terms and conditions for use, reproduction,
      and distribution as defined by Sections 1 through 9 of this document.

      "Licensor" shall mean the copyright owner or entity authorized by
      the copyright owner that is granting the License.

      "Legal Entity" shall mean the union of the acting entity and all
      other entities that control, are controlled by, or are under common
      control with that entity. For the purposes of this definition,
      "control" means (i) the power, direct or indirect, to cause the
      direction or management of such entity, whether by contract or
      otherwise, or (ii) ownership of fifty percent (50%) or more of the
      outstanding shares, or (iii) beneficial ownership of such entity.

      "You" (or "Your") shall mean an individual or Legal Entity
      exercising permissions granted by this License.

      "Source" form shall mean the preferred form for making modifications,
      including but not limited to software source code, documentation
      source, and configuration files.

      "Object" form shall mean any form resulting from mechanical
      transformation or translation of a Source form, including but
      not limited to compiled object code, generated documentation,
      and conversions to other media types.

      "Work" shall mean the work of authorship, whether in Source or
      Object form, made available under the License, as indicated by a
      copyright notice that is included in or attached to the work
      (an example is provided in the Appendix below).

      "Derivative Works" shall mean any work, whether in Source or Object
      form, that is based on (or derived from) the Work and for which the
      editorial revisions, annotations, elaborations, or other modifications
      represent, as a whole, an original work of authorship. For the purposes
      of this License, Derivative Works shall not include works that remain
      separable from, or merely link (or bind by name) to the interfaces of,
      the Work and Derivative Works thereof.

      "Contribution" shall mean any work of authorship, including
      the original version of the Work and any modifications or additions
      to that Work or Derivative Works thereof, that is intentionally
      submitted to Licensor for inclusion in the Work by the copyright owner
      or by an individual or Legal Entity authorized to submit on behalf of
      the copyright owner. For the purposes of this definition, "submitted"
      means any form of electronic, verbal, or written communication sent
      to the Licensor or its representatives, including but not limited to
      communication on electronic mailing lists, source code control systems,
      and issue tracking systems that are managed by, or on behalf of, the
      Licensor for the purpose of discussing and improving the Work, but
      excluding communication that is conspicuously marked or otherwise
      designated in writing by the copyright owner as "Not a Contribution."

      "Contributor" shall mean Licensor and any individual or Legal Entity
      on behalf of whom a Contribution has been received by Licensor and
      subsequently incorporated within the Work.

   2. Grant of Copyright License. Subject to the terms and conditions of
      this License, each Contributor hereby grants to You a perpetual,
      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
      copyright license to reproduce, prepare Derivative Works of,
      publicly display, publicly perform, sublicense, and distribute the
      Work and such Derivative Works in Source or Object form.

   3. Grant of Patent License. Subject to the terms and conditions of
      this License, each Contributor hereby grants to You a perpetual,
      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
      (except as stated in this section) patent license to make, have made,
      use, offer to sell, sell, import, and otherwise transfer the Work,
      where such license applies only to those patent claims licensable
      by such Contributor that are necessarily infringed by their
      Contribution(s) alone or by combination of their Contribution(s)
      with the Work to which such Contribution(s) was submitted. If You
      institute patent litigation against any entity (including a
      cross-claim or counterclaim in a lawsuit) alleging that the Work
      or a Contribution incorporated within the Work constitutes direct
      or contributory patent infringement, then any patent licenses
      granted to You under this License for that Work shall terminate
      as of the date such litigation is filed.

   4. Redistribution. You may reproduce and distribute copies of the
      Work or Derivative Works thereof in any medium, with or without
      modifications, and in Source or Object form, provided that You
      meet the following conditions:

      (a) You must give any other recipients of the Work or
          Derivative Works a copy of this License; and

      (b) You must cause any modified files to carry prominent notices
          stating that You changed the files; and

      (c) You must retain, in the Source form of any Derivative Works
          that You distribute, all copyright, patent, trademark, and
          attribution notices from the Source form of the Work,
          excluding those notices that do not pertain to any part of
          the Derivative Works; and

      (d) If the Work includes a "NOTICE" text file as part of its
          distribution, then any Derivative Works that You distribute must
          include a readable copy of the attribution notices contained
          within such NOTICE file, excluding those notices that do not
          pertain to any part of the Derivative Works, in at least one
          of the following places: within a NOTICE text file distributed
          as part of the Derivative Works; within the Source form or
          documentation, if provided along with the Derivative Works; or,
          within a display generated by the Derivative Works, if and
          wherever such third-party notices normally appear. The contents
          of the NOTICE file are for informational purposes only and
          do not modify the License. You may add Your own attribution
          notices within Derivative Works that You distribute, alongside
          or as an addendum to the NOTICE text from the Work, provided
          that such additional attribution notices cannot be construed
          as modifying the License.

      You may add Your own copyright statement to Your modifications and
      may provide additional or different license terms and conditions
      for use, reproduction, or distribution of Your modifications, or
      for any such Derivative Works as a whole, provided Your use,
      reproduction, and distribution of the Work otherwise complies with
      the conditions stated in this License.

   5. Submission of Contributions. Unless You explicitly state otherwise,
      any Contribution intentionally submitted for inclusion in the Work
      by You to the Licensor shall be under the terms and conditions of
      this License, without any additional terms or conditions.
      Notwithstanding the above, nothing herein shall supersede or modify
      the terms of any separate license agreement you may have executed
      with Licensor regarding such Contributions.

   6. Trademarks. This License does not grant permission to use the trade
      names, trademarks, service marks, or product names of the Licensor,
      except as required for reasonable and customary use in describing the
      origin of the Work and reproducing the content of the NOTICE file.

   7. Disclaimer of Warranty. Unless required by applicable law or
      agreed to in writing, Licensor provides the Work (and each
      Contributor provides its Contributions) on an "AS IS" BASIS,
      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
      implied, including, without limitation, any warranties or conditions
      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
      PARTICULAR PURPOSE. You are solely responsible for determining the
      appropriateness of using or redistributing the Work and assume any
      risks associated with Your exercise of permissions under this License.

   8. Limitation of Liability. In no event and under no legal theory,
      whether in tort (including negligence), contract, or otherwise,
      unless required by applicable law (such as deliberate and grossly
      negligent acts) or agreed to in writing, shall any Contributor be
      liable to You for damages, including any direct, indirect, special,
      incidental, or consequential damages of any character arising as a
      result of this License or out of the use or inability to use the
      Work (including but not limited to damages for loss of goodwill,
      work stoppage, computer failure or malfunction, or any and all
      other commercial damages or losses), even if such Contributor
      has been advised of the possibility of such damages.

   9. Accepting Warranty or Additional Liability. While redistributing
      the Work or Derivative Works thereof, You may choose to offer,
      and charge a fee for, acceptance of support, warranty, indemnity,
      or other liability obligations and/or rights consistent with this
      License. However, in accepting such obligations, You may act only
      on Your own behalf and on Your sole responsibility, not on behalf
      of any other Contributor, and only if You agree to indemnify,
      defend, and hold each Contributor harmless for any liability
      incurred by, or claims asserted against, such Contributor by reason
      of your accepting any such warranty or additional liability.

   END OF TERMS AND CONDITIONS

   APPENDIX: How to apply the Apache License to your work.

      To apply the Apache License to your work, attach the following
      boilerplate notice, with the fields enclosed by brackets "[]"
      replaced with your own identifying information. (Don't include
      the brackets!)  The text should be enclosed in the appropriate
      comment syntax for the file format. We also recommend that a
      file or class name and description of purpose be included on the
      same "printed page" as the copyright notice for easier
      identification within third-party archives.

   Copyright [yyyy] [name of copyright owner]

   Licensed under the Apache License, Version 2.0 (the "License");
   you may not use this file except in compliance with the License.
   You may obtain a copy of the License at

       http://www.apache.org/licenses/LICENSE-2.0

   Unless required by applicable law or agreed to in writing, software
   distributed under the License is distributed on an "AS IS" BASIS,
   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
   See the License for the specific language governing permissions and
   limitations under the License.


================================================
FILE: README.md
================================================
<p align="center">
    <a href="#readme">
        <img alt="logo" width="50%" src="output/evolution-strategy.png">
    </a>
</p>
<p align="center">
  <a href="https://github.com/huseinzol05/Stock-Prediction-Models/blob/master/LICENSE"><img alt="MIT License" src="https://img.shields.io/badge/License-Apache--License--2.0-yellow.svg"></a>
  <a href="#"><img src="https://img.shields.io/badge/deeplearning-30--models-success.svg"></a>
  <a href="#"><img src="https://img.shields.io/badge/agent-23--models-success.svg"></a>
</p>

---

**Stock-Prediction-Models**, Gathers machine learning and deep learning models for Stock forecasting, included trading bots and simulations.

## Table of contents
  * [Models](#models)
  * [Agents](#agents)
  * [Realtime Agent](realtime-agent)
  * [Data Explorations](#data-explorations)
  * [Simulations](#simulations)
  * [Tensorflow-js](#tensorflow-js)
  * [Misc](#misc)
  * [Results](#results)
    * [Results Agent](#results-agent)
    * [Results signal prediction](#results-signal-prediction)
    * [Results analysis](#results-analysis)
    * [Results simulation](#results-simulation)

## Contents

### Models

#### [Deep-learning models](deep-learning)
 1. LSTM
 2. LSTM Bidirectional
 3. LSTM 2-Path
 4. GRU
 5. GRU Bidirectional
 6. GRU 2-Path
 7. Vanilla
 8. Vanilla Bidirectional
 9. Vanilla 2-Path
 10. LSTM Seq2seq
 11. LSTM Bidirectional Seq2seq
 12. LSTM Seq2seq VAE
 13. GRU Seq2seq
 14. GRU Bidirectional Seq2seq
 15. GRU Seq2seq VAE
 16. Attention-is-all-you-Need
 17. CNN-Seq2seq
 18. Dilated-CNN-Seq2seq

**Bonus**

1. How to use one of the model to forecast `t + N`, [how-to-forecast.ipynb](deep-learning/how-to-forecast.ipynb)
2. Consensus, how to use sentiment data to forecast `t + N`, [sentiment-consensus.ipynb](deep-learning/sentiment-consensus.ipynb)

#### [Stacking models](stacking)
 1. Deep Feed-forward Auto-Encoder Neural Network to reduce dimension + Deep Recurrent Neural Network + ARIMA + Extreme Boosting Gradient Regressor
 2. Adaboost + Bagging + Extra Trees + Gradient Boosting + Random Forest + XGB

### [Agents](agent)

1. Turtle-trading agent
2. Moving-average agent
3. Signal rolling agent
4. Policy-gradient agent
5. Q-learning agent
6. Evolution-strategy agent
7. Double Q-learning agent
8. Recurrent Q-learning agent
9. Double Recurrent Q-learning agent
10. Duel Q-learning agent
11. Double Duel Q-learning agent
12. Duel Recurrent Q-learning agent
13. Double Duel Recurrent Q-learning agent
14. Actor-critic agent
15. Actor-critic Duel agent
16. Actor-critic Recurrent agent
17. Actor-critic Duel Recurrent agent
18. Curiosity Q-learning agent
19. Recurrent Curiosity Q-learning agent
20. Duel Curiosity Q-learning agent
21. Neuro-evolution agent
22. Neuro-evolution with Novelty search agent
23. ABCD strategy agent

### [Data Explorations](misc)

1. stock market study on TESLA stock, [tesla-study.ipynb](misc/tesla-study.ipynb)
2. Outliers study using K-means, SVM, and Gaussian on TESLA stock, [outliers.ipynb](misc/outliers.ipynb)
3. Overbought-Oversold study on TESLA stock, [overbought-oversold.ipynb](misc/overbought-oversold.ipynb)
4. Which stock you need to buy? [which-stock.ipynb](misc/which-stock.ipynb)

### [Simulations](simulation)

1. Simple Monte Carlo, [monte-carlo-drift.ipynb](simulation/monte-carlo-drift.ipynb)
2. Dynamic volatility Monte Carlo, [monte-carlo-dynamic-volatility.ipynb](simulation/monte-carlo-dynamic-volatility.ipynb)
3. Drift Monte Carlo, [monte-carlo-drift.ipynb](simulation/monte-carlo-drift.ipynb)
4. Multivariate Drift Monte Carlo BTC/USDT with Bitcurate sentiment, [multivariate-drift-monte-carlo.ipynb](simulation/multivariate-drift-monte-carlo.ipynb)
5. Portfolio optimization, [portfolio-optimization.ipynb](simulation/portfolio-optimization.ipynb), inspired from https://pythonforfinance.net/2017/01/21/investment-portfolio-optimisation-with-python/

### [Tensorflow-js](stock-forecasting-js)

I code [LSTM Recurrent Neural Network](deep-learning/1.lstm.ipynb) and [Simple signal rolling agent](agent/simple-agent.ipynb) inside Tensorflow JS, you can try it here, [huseinhouse.com/stock-forecasting-js](https://huseinhouse.com/stock-forecasting-js/), you can download any historical CSV and upload dynamically.

### [Misc](misc)

1. fashion trending prediction with cross-validation, [fashion-forecasting.ipynb](misc/fashion-forecasting.ipynb)
2. Bitcoin analysis with LSTM prediction, [bitcoin-analysis-lstm.ipynb](misc/bitcoin-analysis-lstm.ipynb)
3. Kijang Emas Bank Negara, [kijang-emas-bank-negara.ipynb](misc/kijang-emas-bank-negara.ipynb)

## Results

### Results Agent

**This agent only able to buy or sell 1 unit per transaction.**

1. Turtle-trading agent, [turtle-agent.ipynb](agent/1.turtle-agent.ipynb)

<img src="output-agent/turtle-agent.png" width="70%" align="">

2. Moving-average agent, [moving-average-agent.ipynb](agent/2.moving-average-agent.ipynb)

<img src="output-agent/moving-average-agent.png" width="70%" align="">

3. Signal rolling agent, [signal-rolling-agent.ipynb](agent/3.signal-rolling-agent.ipynb)

<img src="output-agent/signal-rolling-agent.png" width="70%" align="">

4. Policy-gradient agent, [policy-gradient-agent.ipynb](agent/4.policy-gradient-agent.ipynb)

<img src="output-agent/policy-gradient-agent.png" width="70%" align="">

5. Q-learning agent, [q-learning-agent.ipynb](agent/5.q-learning-agent.ipynb)

<img src="output-agent/q-learning-agent.png" width="70%" align="">

6. Evolution-strategy agent, [evolution-strategy-agent.ipynb](agent/6.evolution-strategy-agent.ipynb)

<img src="output-agent/evolution-strategy-agent.png" width="70%" align="">

7. Double Q-learning agent, [double-q-learning-agent.ipynb](agent/7.double-q-learning-agent.ipynb)

<img src="output-agent/double-q-learning.png" width="70%" align="">

8. Recurrent Q-learning agent, [recurrent-q-learning-agent.ipynb](agent/8.recurrent-q-learning-agent.ipynb)

<img src="output-agent/recurrent-q-learning.png" width="70%" align="">

9. Double Recurrent Q-learning agent, [double-recurrent-q-learning-agent.ipynb](agent/9.double-recurrent-q-learning-agent.ipynb)

<img src="output-agent/double-recurrent-q-learning.png" width="70%" align="">

10. Duel Q-learning agent, [duel-q-learning-agent.ipynb](agent/10.duel-q-learning-agent.ipynb)

<img src="output-agent/double-q-learning.png" width="70%" align="">

11. Double Duel Q-learning agent, [double-duel-q-learning-agent.ipynb](agent/11.double-duel-q-learning-agent.ipynb)

<img src="output-agent/double-duel-q-learning.png" width="70%" align="">

12. Duel Recurrent Q-learning agent, [duel-recurrent-q-learning-agent.ipynb](agent/12.duel-recurrent-q-learning-agent.ipynb)

<img src="output-agent/duel-recurrent-q-learning.png" width="70%" align="">

13. Double Duel Recurrent Q-learning agent, [double-duel-recurrent-q-learning-agent.ipynb](agent/13.double-duel-recurrent-q-learning-agent.ipynb)

<img src="output-agent/double-duel-recurrent-q-learning.png" width="70%" align="">

14. Actor-critic agent, [actor-critic-agent.ipynb](agent/14.actor-critic-agent.ipynb)

<img src="output-agent/actor-critic.png" width="70%" align="">

15. Actor-critic Duel agent, [actor-critic-duel-agent.ipynb](agent/14.actor-critic-duel-agent.ipynb)

<img src="output-agent/actor-critic-duel.png" width="70%" align="">

16. Actor-critic Recurrent agent, [actor-critic-recurrent-agent.ipynb](agent/16.actor-critic-recurrent-agent.ipynb)

<img src="output-agent/actor-critic-recurrent.png" width="70%" align="">

17. Actor-critic Duel Recurrent agent, [actor-critic-duel-recurrent-agent.ipynb](agent/17.actor-critic-duel-recurrent-agent.ipynb)

<img src="output-agent/actor-critic-duel-recurrent.png" width="70%" align="">

18. Curiosity Q-learning agent, [curiosity-q-learning-agent.ipynb](agent/18.curiosity-q-learning-agent.ipynb)

<img src="output-agent/curiosity-q-learning.png" width="70%" align="">

19. Recurrent Curiosity Q-learning agent, [recurrent-curiosity-q-learning.ipynb](agent/19.recurrent-curiosity-q-learning-agent.ipynb)

<img src="output-agent/recurrent-curiosity-q-learning.png" width="70%" align="">

20. Duel Curiosity Q-learning agent, [duel-curiosity-q-learning-agent.ipynb](agent/20.duel-curiosity-q-learning-agent.ipynb)

<img src="output-agent/duel-curiosity-q-learning.png" width="70%" align="">

21. Neuro-evolution agent, [neuro-evolution.ipynb](agent/21.neuro-evolution-agent.ipynb)

<img src="output-agent/neuro-evolution.png" width="70%" align="">

22. Neuro-evolution with Novelty search agent, [neuro-evolution-novelty-search.ipynb](agent/22.neuro-evolution-novelty-search-agent.ipynb)

<img src="output-agent/neuro-evolution-novelty-search.png" width="70%" align="">

23. ABCD strategy agent, [abcd-strategy.ipynb](agent/23.abcd-strategy-agent.ipynb)

<img src="output-agent/abcd-strategy.png" width="70%" align="">

### Results signal prediction

I will cut the dataset to train and test datasets,

1. Train dataset derived from starting timestamp until last 30 days
2. Test dataset derived from last 30 days until end of the dataset

So we will let the model do forecasting based on last 30 days, and we will going to repeat the experiment for 10 times. You can increase it locally if you want, and tuning parameters will help you by a lot.

1. LSTM, accuracy 95.693%, time taken for 1 epoch 01:09

<img src="output/lstm.png" width="70%" align="">

2. LSTM Bidirectional, accuracy 93.8%, time taken for 1 epoch 01:40

<img src="output/bidirectional-lstm.png" width="70%" align="">

3. LSTM 2-Path, accuracy 94.63%, time taken for 1 epoch 01:39

<img src="output/lstm-2path.png" width="70%" align="">

4. GRU, accuracy 94.63%, time taken for 1 epoch 02:10

<img src="output/gru.png" width="70%" align="">

5. GRU Bidirectional, accuracy 92.5673%, time taken for 1 epoch 01:40

<img src="output/bidirectional-gru.png" width="70%" align="">

6. GRU 2-Path, accuracy 93.2117%, time taken for 1 epoch 01:39

<img src="output/gru-2path.png" width="70%" align="">

7. Vanilla, accuracy 91.4686%, time taken for 1 epoch 00:52

<img src="output/vanilla.png" width="70%" align="">

8. Vanilla Bidirectional, accuracy 88.9927%, time taken for 1 epoch 01:06

<img src="output/bidirectional-vanilla.png" width="70%" align="">

9. Vanilla 2-Path, accuracy 91.5406%, time taken for 1 epoch 01:08

<img src="output/vanilla-2path.png" width="70%" align="">

10. LSTM Seq2seq, accuracy 94.9817%, time taken for 1 epoch 01:36

<img src="output/lstm-seq2seq.png" width="70%" align="">

11. LSTM Bidirectional Seq2seq, accuracy 94.517%, time taken for 1 epoch 02:30

<img src="output/bidirectional-lstm-seq2seq.png" width="70%" align="">

12. LSTM Seq2seq VAE, accuracy 95.4190%, time taken for 1 epoch 01:48

<img src="output/lstm-seq2seq-vae.png" width="70%" align="">

13. GRU Seq2seq, accuracy 90.8854%, time taken for 1 epoch 01:34

<img src="output/gru-seq2seq.png" width="70%" align="">

14. GRU Bidirectional Seq2seq, accuracy 67.9915%, time taken for 1 epoch 02:30

<img src="output/bidirectional-gru-seq2seq.png" width="70%" align="">

15. GRU Seq2seq VAE, accuracy 89.1321%, time taken for 1 epoch 01:48

<img src="output/gru-seq2seq-vae.png" width="70%" align="">

16. Attention-is-all-you-Need, accuracy 94.2482%, time taken for 1 epoch 01:41

<img src="output/attention-is-all-you-need.png" width="70%" align="">

17. CNN-Seq2seq, accuracy 90.74%, time taken for 1 epoch 00:43

<img src="output/cnn-seq2seq.png" width="70%" align="">

18. Dilated-CNN-Seq2seq, accuracy 95.86%, time taken for 1 epoch 00:14

<img src="output/dilated-cnn-seq2seq.png" width="70%" align="">

**Bonus**

1. How to forecast,

<img src="output/how-to-forecast.png" width="70%" align="">

2. Sentiment consensus,

<img src="output/sentiment-consensus.png" width="70%" align="">

### Results analysis

1. Outliers study using K-means, SVM, and Gaussian on TESLA stock

<img src="misc/outliers.png" width="70%" align="">

2. Overbought-Oversold study on TESLA stock

<img src="misc/overbought-oversold.png" width="70%" align="">

3. Which stock you need to buy?

<img src="misc/which-stock.png" width="40%" align="">

### Results simulation

1. Simple Monte Carlo

<img src="simulation/monte-carlo-simple.png" width="70%" align="">

2. Dynamic volatity Monte Carlo

<img src="simulation/monte-carlo-dynamic-volatility.png" width="70%" align="">

3. Drift Monte Carlo

<img src="simulation/monte-carlo-drift.png" width="70%" align="">

4. Multivariate Drift Monte Carlo BTC/USDT with Bitcurate sentiment

<img src="simulation/multivariate-drift-monte-carlo.png" width="70%" align="">

5. Portfolio optimization

<img src="simulation/portfolio-optimization.png" width="40%" align="">


================================================
FILE: agent/1.turtle-agent.ipynb
================================================
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Date</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Adj Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2016-11-02</td>\n",
       "      <td>778.200012</td>\n",
       "      <td>781.650024</td>\n",
       "      <td>763.450012</td>\n",
       "      <td>768.700012</td>\n",
       "      <td>768.700012</td>\n",
       "      <td>1872400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2016-11-03</td>\n",
       "      <td>767.250000</td>\n",
       "      <td>769.950012</td>\n",
       "      <td>759.030029</td>\n",
       "      <td>762.130005</td>\n",
       "      <td>762.130005</td>\n",
       "      <td>1943200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016-11-04</td>\n",
       "      <td>750.659973</td>\n",
       "      <td>770.359985</td>\n",
       "      <td>750.560974</td>\n",
       "      <td>762.020020</td>\n",
       "      <td>762.020020</td>\n",
       "      <td>2134800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016-11-07</td>\n",
       "      <td>774.500000</td>\n",
       "      <td>785.190002</td>\n",
       "      <td>772.549988</td>\n",
       "      <td>782.520020</td>\n",
       "      <td>782.520020</td>\n",
       "      <td>1585100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016-11-08</td>\n",
       "      <td>783.400024</td>\n",
       "      <td>795.632996</td>\n",
       "      <td>780.190002</td>\n",
       "      <td>790.510010</td>\n",
       "      <td>790.510010</td>\n",
       "      <td>1350800</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Date        Open        High         Low       Close   Adj Close  \\\n",
       "0  2016-11-02  778.200012  781.650024  763.450012  768.700012  768.700012   \n",
       "1  2016-11-03  767.250000  769.950012  759.030029  762.130005  762.130005   \n",
       "2  2016-11-04  750.659973  770.359985  750.560974  762.020020  762.020020   \n",
       "3  2016-11-07  774.500000  785.190002  772.549988  782.520020  782.520020   \n",
       "4  2016-11-08  783.400024  795.632996  780.190002  790.510010  790.510010   \n",
       "\n",
       "    Volume  \n",
       "0  1872400  \n",
       "1  1943200  \n",
       "2  2134800  \n",
       "3  1585100  \n",
       "4  1350800  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('../dataset/GOOG-year.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>signal</th>\n",
       "      <th>trend</th>\n",
       "      <th>RollingMax</th>\n",
       "      <th>RollingMin</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>768.700012</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>762.130005</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>762.020020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>782.520020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>790.510010</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.0</td>\n",
       "      <td>785.309998</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.0</td>\n",
       "      <td>762.559998</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.0</td>\n",
       "      <td>754.020020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.0</td>\n",
       "      <td>736.080017</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.0</td>\n",
       "      <td>758.489990</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.0</td>\n",
       "      <td>764.479980</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.0</td>\n",
       "      <td>771.229980</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.0</td>\n",
       "      <td>760.539978</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.0</td>\n",
       "      <td>769.200012</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.0</td>\n",
       "      <td>768.270020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.0</td>\n",
       "      <td>760.989990</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.0</td>\n",
       "      <td>761.679993</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.0</td>\n",
       "      <td>768.239990</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.0</td>\n",
       "      <td>770.840027</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.0</td>\n",
       "      <td>758.039978</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.0</td>\n",
       "      <td>747.919983</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.0</td>\n",
       "      <td>750.500000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.0</td>\n",
       "      <td>762.520020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.0</td>\n",
       "      <td>759.109985</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.0</td>\n",
       "      <td>771.190002</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.0</td>\n",
       "      <td>776.419983</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0.0</td>\n",
       "      <td>789.289978</td>\n",
       "      <td>790.510010</td>\n",
       "      <td>736.080017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.0</td>\n",
       "      <td>789.270020</td>\n",
       "      <td>790.510010</td>\n",
       "      <td>736.080017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>796.099976</td>\n",
       "      <td>790.510010</td>\n",
       "      <td>736.080017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>797.070007</td>\n",
       "      <td>796.099976</td>\n",
       "      <td>736.080017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>222</th>\n",
       "      <td>0.0</td>\n",
       "      <td>932.450012</td>\n",
       "      <td>939.330017</td>\n",
       "      <td>906.659973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>223</th>\n",
       "      <td>0.0</td>\n",
       "      <td>928.530029</td>\n",
       "      <td>939.330017</td>\n",
       "      <td>906.659973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>224</th>\n",
       "      <td>0.0</td>\n",
       "      <td>920.969971</td>\n",
       "      <td>939.330017</td>\n",
       "      <td>906.659973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>225</th>\n",
       "      <td>0.0</td>\n",
       "      <td>924.859985</td>\n",
       "      <td>939.330017</td>\n",
       "      <td>906.659973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>226</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>944.489990</td>\n",
       "      <td>939.330017</td>\n",
       "      <td>906.659973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>227</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>949.500000</td>\n",
       "      <td>944.489990</td>\n",
       "      <td>913.809998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>228</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>959.109985</td>\n",
       "      <td>949.500000</td>\n",
       "      <td>913.809998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>229</th>\n",
       "      <td>0.0</td>\n",
       "      <td>953.270020</td>\n",
       "      <td>959.109985</td>\n",
       "      <td>913.809998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>230</th>\n",
       "      <td>0.0</td>\n",
       "      <td>957.789978</td>\n",
       "      <td>959.109985</td>\n",
       "      <td>913.809998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>231</th>\n",
       "      <td>0.0</td>\n",
       "      <td>951.679993</td>\n",
       "      <td>959.109985</td>\n",
       "      <td>913.809998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>232</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>969.960022</td>\n",
       "      <td>959.109985</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>233</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>978.890015</td>\n",
       "      <td>969.960022</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>234</th>\n",
       "      <td>0.0</td>\n",
       "      <td>977.000000</td>\n",
       "      <td>978.890015</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>235</th>\n",
       "      <td>0.0</td>\n",
       "      <td>972.599976</td>\n",
       "      <td>978.890015</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>236</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>989.250000</td>\n",
       "      <td>978.890015</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>237</th>\n",
       "      <td>0.0</td>\n",
       "      <td>987.830017</td>\n",
       "      <td>989.250000</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>238</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>989.679993</td>\n",
       "      <td>989.250000</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>239</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>992.000000</td>\n",
       "      <td>989.679993</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>240</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>992.179993</td>\n",
       "      <td>992.000000</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>241</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>992.809998</td>\n",
       "      <td>992.179993</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>242</th>\n",
       "      <td>0.0</td>\n",
       "      <td>984.450012</td>\n",
       "      <td>992.809998</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>243</th>\n",
       "      <td>0.0</td>\n",
       "      <td>988.200012</td>\n",
       "      <td>992.809998</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>244</th>\n",
       "      <td>0.0</td>\n",
       "      <td>968.450012</td>\n",
       "      <td>992.809998</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>245</th>\n",
       "      <td>0.0</td>\n",
       "      <td>970.539978</td>\n",
       "      <td>992.809998</td>\n",
       "      <td>915.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>246</th>\n",
       "      <td>0.0</td>\n",
       "      <td>973.330017</td>\n",
       "      <td>992.809998</td>\n",
       "      <td>920.969971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>0.0</td>\n",
       "      <td>972.559998</td>\n",
       "      <td>992.809998</td>\n",
       "      <td>920.969971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>248</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>1019.270020</td>\n",
       "      <td>992.809998</td>\n",
       "      <td>920.969971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>249</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1017.109985</td>\n",
       "      <td>1019.270020</td>\n",
       "      <td>920.969971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>250</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1016.640015</td>\n",
       "      <td>1019.270020</td>\n",
       "      <td>920.969971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>251</th>\n",
       "      <td>-1.0</td>\n",
       "      <td>1025.500000</td>\n",
       "      <td>1019.270020</td>\n",
       "      <td>924.859985</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>252 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     signal        trend   RollingMax  RollingMin\n",
       "0       0.0   768.700012          NaN         NaN\n",
       "1       0.0   762.130005          NaN         NaN\n",
       "2       0.0   762.020020          NaN         NaN\n",
       "3       0.0   782.520020          NaN         NaN\n",
       "4       0.0   790.510010          NaN         NaN\n",
       "5       0.0   785.309998          NaN         NaN\n",
       "6       0.0   762.559998          NaN         NaN\n",
       "7       0.0   754.020020          NaN         NaN\n",
       "8       0.0   736.080017          NaN         NaN\n",
       "9       0.0   758.489990          NaN         NaN\n",
       "10      0.0   764.479980          NaN         NaN\n",
       "11      0.0   771.229980          NaN         NaN\n",
       "12      0.0   760.539978          NaN         NaN\n",
       "13      0.0   769.200012          NaN         NaN\n",
       "14      0.0   768.270020          NaN         NaN\n",
       "15      0.0   760.989990          NaN         NaN\n",
       "16      0.0   761.679993          NaN         NaN\n",
       "17      0.0   768.239990          NaN         NaN\n",
       "18      0.0   770.840027          NaN         NaN\n",
       "19      0.0   758.039978          NaN         NaN\n",
       "20      0.0   747.919983          NaN         NaN\n",
       "21      0.0   750.500000          NaN         NaN\n",
       "22      0.0   762.520020          NaN         NaN\n",
       "23      0.0   759.109985          NaN         NaN\n",
       "24      0.0   771.190002          NaN         NaN\n",
       "25      0.0   776.419983          NaN         NaN\n",
       "26      0.0   789.289978   790.510010  736.080017\n",
       "27      0.0   789.270020   790.510010  736.080017\n",
       "28     -1.0   796.099976   790.510010  736.080017\n",
       "29     -1.0   797.070007   796.099976  736.080017\n",
       "..      ...          ...          ...         ...\n",
       "222     0.0   932.450012   939.330017  906.659973\n",
       "223     0.0   928.530029   939.330017  906.659973\n",
       "224     0.0   920.969971   939.330017  906.659973\n",
       "225     0.0   924.859985   939.330017  906.659973\n",
       "226    -1.0   944.489990   939.330017  906.659973\n",
       "227    -1.0   949.500000   944.489990  913.809998\n",
       "228    -1.0   959.109985   949.500000  913.809998\n",
       "229     0.0   953.270020   959.109985  913.809998\n",
       "230     0.0   957.789978   959.109985  913.809998\n",
       "231     0.0   951.679993   959.109985  913.809998\n",
       "232    -1.0   969.960022   959.109985  915.000000\n",
       "233    -1.0   978.890015   969.960022  915.000000\n",
       "234     0.0   977.000000   978.890015  915.000000\n",
       "235     0.0   972.599976   978.890015  915.000000\n",
       "236    -1.0   989.250000   978.890015  915.000000\n",
       "237     0.0   987.830017   989.250000  915.000000\n",
       "238    -1.0   989.679993   989.250000  915.000000\n",
       "239    -1.0   992.000000   989.679993  915.000000\n",
       "240    -1.0   992.179993   992.000000  915.000000\n",
       "241    -1.0   992.809998   992.179993  915.000000\n",
       "242     0.0   984.450012   992.809998  915.000000\n",
       "243     0.0   988.200012   992.809998  915.000000\n",
       "244     0.0   968.450012   992.809998  915.000000\n",
       "245     0.0   970.539978   992.809998  915.000000\n",
       "246     0.0   973.330017   992.809998  920.969971\n",
       "247     0.0   972.559998   992.809998  920.969971\n",
       "248    -1.0  1019.270020   992.809998  920.969971\n",
       "249     0.0  1017.109985  1019.270020  920.969971\n",
       "250     0.0  1016.640015  1019.270020  920.969971\n",
       "251    -1.0  1025.500000  1019.270020  924.859985\n",
       "\n",
       "[252 rows x 4 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "count = int(np.ceil(len(df) * 0.1))\n",
    "signals = pd.DataFrame(index=df.index)\n",
    "signals['signal'] = 0.0\n",
    "signals['trend'] = df['Close']\n",
    "signals['RollingMax'] = (signals.trend.shift(1).rolling(count).max())\n",
    "signals['RollingMin'] = (signals.trend.shift(1).rolling(count).min())\n",
    "signals.loc[signals['RollingMax'] < signals.trend, 'signal'] = -1\n",
    "signals.loc[signals['RollingMin'] > signals.trend, 'signal'] = 1\n",
    "signals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def buy_stock(\n",
    "    real_movement,\n",
    "    signal,\n",
    "    initial_money = 10000,\n",
    "    max_buy = 1,\n",
    "    max_sell = 1,\n",
    "):\n",
    "    \"\"\"\n",
    "    real_movement = actual movement in the real world\n",
    "    delay = how much interval you want to delay to change our decision from buy to sell, vice versa\n",
    "    initial_state = 1 is buy, 0 is sell\n",
    "    initial_money = 1000, ignore what kind of currency\n",
    "    max_buy = max quantity for share to buy\n",
    "    max_sell = max quantity for share to sell\n",
    "    \"\"\"\n",
    "    starting_money = initial_money\n",
    "    states_sell = []\n",
    "    states_buy = []\n",
    "    current_inventory = 0\n",
    "\n",
    "    def buy(i, initial_money, current_inventory):\n",
    "        shares = initial_money // real_movement[i]\n",
    "        if shares < 1:\n",
    "            print(\n",
    "                'day %d: total balances %f, not enough money to buy a unit price %f'\n",
    "                % (i, initial_money, real_movement[i])\n",
    "            )\n",
    "        else:\n",
    "            if shares > max_buy:\n",
    "                buy_units = max_buy\n",
    "            else:\n",
    "                buy_units = shares\n",
    "            initial_money -= buy_units * real_movement[i]\n",
    "            current_inventory += buy_units\n",
    "            print(\n",
    "                'day %d: buy %d units at price %f, total balance %f'\n",
    "                % (i, buy_units, buy_units * real_movement[i], initial_money)\n",
    "            )\n",
    "            states_buy.append(0)\n",
    "        return initial_money, current_inventory\n",
    "\n",
    "    for i in range(real_movement.shape[0] - int(0.025 * len(df))):\n",
    "        state = signal[i]\n",
    "        if state == 1:\n",
    "            initial_money, current_inventory = buy(\n",
    "                i, initial_money, current_inventory\n",
    "            )\n",
    "            states_buy.append(i)\n",
    "        elif state == -1:\n",
    "            if current_inventory == 0:\n",
    "                    print('day %d: cannot sell anything, inventory 0' % (i))\n",
    "            else:\n",
    "                if current_inventory > max_sell:\n",
    "                    sell_units = max_sell\n",
    "                else:\n",
    "                    sell_units = current_inventory\n",
    "                current_inventory -= sell_units\n",
    "                total_sell = sell_units * real_movement[i]\n",
    "                initial_money += total_sell\n",
    "                try:\n",
    "                    invest = (\n",
    "                        (real_movement[i] - real_movement[states_buy[-1]])\n",
    "                        / real_movement[states_buy[-1]]\n",
    "                    ) * 100\n",
    "                except:\n",
    "                    invest = 0\n",
    "                print(\n",
    "                    'day %d, sell %d units at price %f, investment %f %%, total balance %f,'\n",
    "                    % (i, sell_units, total_sell, invest, initial_money)\n",
    "                )\n",
    "            states_sell.append(i)\n",
    "            \n",
    "    invest = ((initial_money - starting_money) / starting_money) * 100\n",
    "    total_gains = initial_money - starting_money\n",
    "    return states_buy, states_sell, total_gains, invest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "day 28: cannot sell anything, inventory 0\n",
      "day 29: cannot sell anything, inventory 0\n",
      "day 30: cannot sell anything, inventory 0\n",
      "day 44: cannot sell anything, inventory 0\n",
      "day 45: cannot sell anything, inventory 0\n",
      "day 47: cannot sell anything, inventory 0\n",
      "day 54: cannot sell anything, inventory 0\n",
      "day 55: cannot sell anything, inventory 0\n",
      "day 56: cannot sell anything, inventory 0\n",
      "day 85: cannot sell anything, inventory 0\n",
      "day 86: cannot sell anything, inventory 0\n",
      "day 87: cannot sell anything, inventory 0\n",
      "day 88: cannot sell anything, inventory 0\n",
      "day 89: cannot sell anything, inventory 0\n",
      "day 90: cannot sell anything, inventory 0\n",
      "day 91: cannot sell anything, inventory 0\n",
      "day 92: cannot sell anything, inventory 0\n",
      "day 96: buy 1 units at price 817.580017, total balance 9182.419983\n",
      "day 97: buy 1 units at price 814.429993, total balance 8367.989990\n",
      "day 117, sell 1 units at price 862.760010, investment 5.934214 %, total balance 9230.750000,\n",
      "day 118, sell 1 units at price 872.299988, investment 7.105582 %, total balance 10103.049988,\n",
      "day 120: cannot sell anything, inventory 0\n",
      "day 121: cannot sell anything, inventory 0\n",
      "day 122: cannot sell anything, inventory 0\n",
      "day 123: cannot sell anything, inventory 0\n",
      "day 124: cannot sell anything, inventory 0\n",
      "day 125: cannot sell anything, inventory 0\n",
      "day 127: cannot sell anything, inventory 0\n",
      "day 132: cannot sell anything, inventory 0\n",
      "day 133: cannot sell anything, inventory 0\n",
      "day 138: cannot sell anything, inventory 0\n",
      "day 139: cannot sell anything, inventory 0\n",
      "day 140: cannot sell anything, inventory 0\n",
      "day 141: cannot sell anything, inventory 0\n",
      "day 142: cannot sell anything, inventory 0\n",
      "day 146: cannot sell anything, inventory 0\n",
      "day 162: buy 1 units at price 927.330017, total balance 9175.719971\n",
      "day 164: buy 1 units at price 917.789978, total balance 8257.929993\n",
      "day 165: buy 1 units at price 908.729980, total balance 7349.200013\n",
      "day 166: buy 1 units at price 898.700012, total balance 6450.500001\n",
      "day 177, sell 1 units at price 970.890015, investment 8.032714 %, total balance 7421.390016,\n",
      "day 179, sell 1 units at price 972.919983, investment 8.258592 %, total balance 8394.309999,\n",
      "day 180, sell 1 units at price 980.340027, investment 9.084234 %, total balance 9374.650026,\n",
      "day 200: buy 1 units at price 906.659973, total balance 8467.990053\n",
      "day 226, sell 1 units at price 944.489990, investment 4.172459 %, total balance 9412.480043,\n",
      "day 227, sell 1 units at price 949.500000, investment 4.725038 %, total balance 10361.980043,\n",
      "day 228: cannot sell anything, inventory 0\n",
      "day 232: cannot sell anything, inventory 0\n",
      "day 233: cannot sell anything, inventory 0\n",
      "day 236: cannot sell anything, inventory 0\n",
      "day 238: cannot sell anything, inventory 0\n",
      "day 239: cannot sell anything, inventory 0\n",
      "day 240: cannot sell anything, inventory 0\n",
      "day 241: cannot sell anything, inventory 0\n"
     ]
    }
   ],
   "source": [
    "states_buy, states_sell, total_gains, invest = buy_stock(df.Close, signals['signal'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "close = df['Close']\n",
    "fig = plt.figure(figsize = (15,5))\n",
    "plt.plot(close, color='r', lw=2.)\n",
    "plt.plot(close, '^', markersize=10, color='m', label = 'buying signal', markevery = states_buy)\n",
    "plt.plot(close, 'v', markersize=10, color='k', label = 'selling signal', markevery = states_sell)\n",
    "plt.title('total gains %f, total investment %f%%'%(total_gains, invest))\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}


================================================
FILE: agent/10.duel-q-learning-agent.ipynb
================================================
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Date</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Adj Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2016-11-02</td>\n",
       "      <td>778.200012</td>\n",
       "      <td>781.650024</td>\n",
       "      <td>763.450012</td>\n",
       "      <td>768.700012</td>\n",
       "      <td>768.700012</td>\n",
       "      <td>1872400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2016-11-03</td>\n",
       "      <td>767.250000</td>\n",
       "      <td>769.950012</td>\n",
       "      <td>759.030029</td>\n",
       "      <td>762.130005</td>\n",
       "      <td>762.130005</td>\n",
       "      <td>1943200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016-11-04</td>\n",
       "      <td>750.659973</td>\n",
       "      <td>770.359985</td>\n",
       "      <td>750.560974</td>\n",
       "      <td>762.020020</td>\n",
       "      <td>762.020020</td>\n",
       "      <td>2134800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016-11-07</td>\n",
       "      <td>774.500000</td>\n",
       "      <td>785.190002</td>\n",
       "      <td>772.549988</td>\n",
       "      <td>782.520020</td>\n",
       "      <td>782.520020</td>\n",
       "      <td>1585100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016-11-08</td>\n",
       "      <td>783.400024</td>\n",
       "      <td>795.632996</td>\n",
       "      <td>780.190002</td>\n",
       "      <td>790.510010</td>\n",
       "      <td>790.510010</td>\n",
       "      <td>1350800</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Date        Open        High         Low       Close   Adj Close  \\\n",
       "0  2016-11-02  778.200012  781.650024  763.450012  768.700012  768.700012   \n",
       "1  2016-11-03  767.250000  769.950012  759.030029  762.130005  762.130005   \n",
       "2  2016-11-04  750.659973  770.359985  750.560974  762.020020  762.020020   \n",
       "3  2016-11-07  774.500000  785.190002  772.549988  782.520020  782.520020   \n",
       "4  2016-11-08  783.400024  795.632996  780.190002  790.510010  790.510010   \n",
       "\n",
       "    Volume  \n",
       "0  1872400  \n",
       "1  1943200  \n",
       "2  2134800  \n",
       "3  1585100  \n",
       "4  1350800  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('../dataset/GOOG-year.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import deque\n",
    "import random\n",
    "\n",
    "\n",
    "class Agent:\n",
    "    def __init__(self, state_size, window_size, trend, skip, batch_size):\n",
    "        self.state_size = state_size\n",
    "        self.window_size = window_size\n",
    "        self.half_window = window_size // 2\n",
    "        self.trend = trend\n",
    "        self.skip = skip\n",
    "        self.action_size = 3\n",
    "        self.batch_size = batch_size\n",
    "        self.memory = deque(maxlen = 1000)\n",
    "        self.inventory = []\n",
    "\n",
    "        self.gamma = 0.95\n",
    "        self.epsilon = 0.5\n",
    "        self.epsilon_min = 0.01\n",
    "        self.epsilon_decay = 0.999\n",
    "\n",
    "        tf.reset_default_graph()\n",
    "        self.sess = tf.InteractiveSession()\n",
    "        self.X = tf.placeholder(tf.float32, [None, self.state_size])\n",
    "        self.Y = tf.placeholder(tf.float32, [None, self.action_size])\n",
    "        feed = tf.layers.dense(self.X, 512, activation = tf.nn.relu)\n",
    "        tensor_action, tensor_validation = tf.split(feed,2,1)\n",
    "        feed_action = tf.layers.dense(tensor_action, self.action_size)\n",
    "        feed_validation = tf.layers.dense(tensor_validation, 1)\n",
    "        self.logits = feed_validation + tf.subtract(feed_action,tf.reduce_mean(feed_action,axis=1,keep_dims=True))\n",
    "        self.cost = tf.reduce_mean(tf.square(self.Y - self.logits))\n",
    "        self.optimizer = tf.train.GradientDescentOptimizer(1e-5).minimize(\n",
    "            self.cost\n",
    "        )\n",
    "        self.sess.run(tf.global_variables_initializer())\n",
    "\n",
    "    def act(self, state):\n",
    "        if random.random() <= self.epsilon:\n",
    "            return random.randrange(self.action_size)\n",
    "        return np.argmax(\n",
    "            self.sess.run(self.logits, feed_dict = {self.X: state})[0]\n",
    "        )\n",
    "    \n",
    "    def get_state(self, t):\n",
    "        window_size = self.window_size + 1\n",
    "        d = t - window_size + 1\n",
    "        block = self.trend[d : t + 1] if d >= 0 else -d * [self.trend[0]] + self.trend[0 : t + 1]\n",
    "        res = []\n",
    "        for i in range(window_size - 1):\n",
    "            res.append(block[i + 1] - block[i])\n",
    "        return np.array([res])\n",
    "\n",
    "    def replay(self, batch_size):\n",
    "        mini_batch = []\n",
    "        l = len(self.memory)\n",
    "        for i in range(l - batch_size, l):\n",
    "            mini_batch.append(self.memory[i])\n",
    "        replay_size = len(mini_batch)\n",
    "        X = np.empty((replay_size, self.state_size))\n",
    "        Y = np.empty((replay_size, self.action_size))\n",
    "        states = np.array([a[0][0] for a in mini_batch])\n",
    "        new_states = np.array([a[3][0] for a in mini_batch])\n",
    "        Q = self.sess.run(self.logits, feed_dict = {self.X: states})\n",
    "        Q_new = self.sess.run(self.logits, feed_dict = {self.X: new_states})\n",
    "        for i in range(len(mini_batch)):\n",
    "            state, action, reward, next_state, done = mini_batch[i]\n",
    "            target = Q[i]\n",
    "            target[action] = reward\n",
    "            if not done:\n",
    "                target[action] += self.gamma * np.amax(Q_new[i])\n",
    "            X[i] = state\n",
    "            Y[i] = target\n",
    "        cost, _ = self.sess.run(\n",
    "            [self.cost, self.optimizer], feed_dict = {self.X: X, self.Y: Y}\n",
    "        )\n",
    "        if self.epsilon > self.epsilon_min:\n",
    "            self.epsilon *= self.epsilon_decay\n",
    "        return cost\n",
    "    \n",
    "    def buy(self, initial_money):\n",
    "        starting_money = initial_money\n",
    "        states_sell = []\n",
    "        states_buy = []\n",
    "        inventory = []\n",
    "        state = self.get_state(0)\n",
    "        for t in range(0, len(self.trend) - 1, self.skip):\n",
    "            action = self.act(state)\n",
    "            next_state = self.get_state(t + 1)\n",
    "            \n",
    "            if action == 1 and initial_money >= self.trend[t] and t < (len(self.trend) - self.half_window):\n",
    "                inventory.append(self.trend[t])\n",
    "                initial_money -= self.trend[t]\n",
    "                states_buy.append(t)\n",
    "                print('day %d: buy 1 unit at price %f, total balance %f'% (t, self.trend[t], initial_money))\n",
    "                \n",
    "                \n",
    "            elif action == 2 and len(inventory):\n",
    "                bought_price = inventory.pop(0)\n",
    "                initial_money += self.trend[t]\n",
    "                states_sell.append(t)\n",
    "                try:\n",
    "                    invest = ((close[t] - bought_price) / bought_price) * 100\n",
    "                except:\n",
    "                    invest = 0\n",
    "                print(\n",
    "                    'day %d, sell 1 unit at price %f, investment %f %%, total balance %f,'\n",
    "                    % (t, close[t], invest, initial_money)\n",
    "                )\n",
    "            \n",
    "            state = next_state\n",
    "        invest = ((initial_money - starting_money) / starting_money) * 100\n",
    "        total_gains = initial_money - starting_money\n",
    "        return states_buy, states_sell, total_gains, invest\n",
    "        \n",
    "    def train(self, iterations, checkpoint, initial_money):\n",
    "        for i in range(iterations):\n",
    "            total_profit = 0\n",
    "            inventory = []\n",
    "            state = self.get_state(0)\n",
    "            starting_money = initial_money\n",
    "            for t in range(0, len(self.trend) - 1, self.skip):\n",
    "                action = self.act(state)\n",
    "                next_state = self.get_state(t + 1)\n",
    "                \n",
    "                if action == 1 and starting_money >= self.trend[t] and t < (len(self.trend) - self.half_window):\n",
    "                    inventory.append(self.trend[t])\n",
    "                    starting_money -= self.trend[t]\n",
    "                \n",
    "                elif action == 2 and len(inventory) > 0:\n",
    "                    bought_price = inventory.pop(0)\n",
    "                    total_profit += self.trend[t] - bought_price\n",
    "                    starting_money += self.trend[t]\n",
    "                    \n",
    "                invest = ((starting_money - initial_money) / initial_money)\n",
    "                self.memory.append((state, action, invest, \n",
    "                                    next_state, starting_money < initial_money))\n",
    "                state = next_state\n",
    "                batch_size = min(self.batch_size, len(self.memory))\n",
    "                cost = self.replay(batch_size)\n",
    "            if (i+1) % checkpoint == 0:\n",
    "                print('epoch: %d, total rewards: %f.3, cost: %f, total money: %f'%(i + 1, total_profit, cost,\n",
    "                                                                                  starting_money))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From <ipython-input-3-28bed545c0f8>:30: calling reduce_mean (from tensorflow.python.ops.math_ops) with keep_dims is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "keep_dims is deprecated, use keepdims instead\n",
      "epoch: 10, total rewards: 231.100222.3, cost: 0.499693, total money: 10231.100222\n",
      "epoch: 20, total rewards: 195.875063.3, cost: 0.324152, total money: 10195.875063\n",
      "epoch: 30, total rewards: 219.615054.3, cost: 0.237771, total money: 10219.615054\n",
      "epoch: 40, total rewards: 56.505131.3, cost: 0.183305, total money: 10056.505131\n",
      "epoch: 50, total rewards: 190.745120.3, cost: 0.129967, total money: 10190.745120\n",
      "epoch: 60, total rewards: 165.275088.3, cost: 0.134246, total money: 10165.275088\n",
      "epoch: 70, total rewards: 201.795107.3, cost: 0.075016, total money: 10201.795107\n",
      "epoch: 80, total rewards: 187.545045.3, cost: 0.062454, total money: 10187.545045\n",
      "epoch: 90, total rewards: 206.835023.3, cost: 0.050687, total money: 10206.835023\n",
      "epoch: 100, total rewards: 199.895082.3, cost: 0.041359, total money: 10199.895082\n",
      "epoch: 110, total rewards: 184.405092.3, cost: 0.035289, total money: 10184.405092\n",
      "epoch: 120, total rewards: 242.405092.3, cost: 0.047248, total money: 10242.405092\n",
      "epoch: 130, total rewards: 148.405032.3, cost: 0.050786, total money: 10148.405032\n",
      "epoch: 140, total rewards: 225.724978.3, cost: 0.021171, total money: 10225.724978\n",
      "epoch: 150, total rewards: 168.344972.3, cost: 0.018388, total money: 10168.344972\n",
      "epoch: 160, total rewards: 230.095034.3, cost: 0.199324, total money: 10230.095034\n",
      "epoch: 170, total rewards: 206.275026.3, cost: 0.044696, total money: 10206.275026\n",
      "epoch: 180, total rewards: 364.895023.3, cost: 0.016494, total money: 10364.895023\n",
      "epoch: 190, total rewards: 220.664980.3, cost: 0.014381, total money: 10220.664980\n",
      "epoch: 200, total rewards: 175.284975.3, cost: 0.010883, total money: 10175.284975\n"
     ]
    }
   ],
   "source": [
    "close = df.Close.values.tolist()\n",
    "initial_money = 10000\n",
    "window_size = 30\n",
    "skip = 1\n",
    "batch_size = 32\n",
    "agent = Agent(state_size = window_size, \n",
    "              window_size = window_size, \n",
    "              trend = close, \n",
    "              skip = skip, \n",
    "              batch_size = batch_size)\n",
    "agent.train(iterations = 200, checkpoint = 10, initial_money = initial_money)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "day 7: buy 1 unit at price 754.020020, total balance 9245.979980\n",
      "day 8: buy 1 unit at price 736.080017, total balance 8509.899963\n",
      "day 9, sell 1 unit at price 758.489990, investment 0.592818 %, total balance 9268.389953,\n",
      "day 11: buy 1 unit at price 771.229980, total balance 8497.159973\n",
      "day 12, sell 1 unit at price 760.539978, investment 3.323003 %, total balance 9257.699951,\n",
      "day 14, sell 1 unit at price 768.270020, investment -0.383797 %, total balance 10025.969971,\n",
      "day 22: buy 1 unit at price 762.520020, total balance 9263.449951\n",
      "day 24, sell 1 unit at price 771.190002, investment 1.137017 %, total balance 10034.639953,\n",
      "day 28: buy 1 unit at price 796.099976, total balance 9238.539977\n",
      "day 29: buy 1 unit at price 797.070007, total balance 8441.469970\n",
      "day 31, sell 1 unit at price 790.799988, investment -0.665744 %, total balance 9232.269958,\n",
      "day 32: buy 1 unit at price 794.200012, total balance 8438.069946\n",
      "day 33: buy 1 unit at price 796.419983, total balance 7641.649963\n",
      "day 37, sell 1 unit at price 791.549988, investment -0.692539 %, total balance 8433.199951,\n",
      "day 39, sell 1 unit at price 782.789978, investment -1.436670 %, total balance 9215.989929,\n",
      "day 41, sell 1 unit at price 786.140015, investment -1.290772 %, total balance 10002.129944,\n",
      "day 48: buy 1 unit at price 806.359985, total balance 9195.769959\n",
      "day 49: buy 1 unit at price 807.880005, total balance 8387.889954\n",
      "day 50, sell 1 unit at price 804.609985, investment -0.217025 %, total balance 9192.499939,\n",
      "day 52: buy 1 unit at price 802.174988, total balance 8390.324951\n",
      "day 53, sell 1 unit at price 805.020020, investment -0.354011 %, total balance 9195.344971,\n",
      "day 56, sell 1 unit at price 835.669983, investment 4.175522 %, total balance 10031.014954,\n",
      "day 67: buy 1 unit at price 809.559998, total balance 9221.454956\n",
      "day 68: buy 1 unit at price 813.669983, total balance 8407.784973\n",
      "day 69, sell 1 unit at price 819.239990, investment 1.195710 %, total balance 9227.024963,\n",
      "day 71: buy 1 unit at price 818.979980, total balance 8408.044983\n",
      "day 72, sell 1 unit at price 824.159973, investment 1.289219 %, total balance 9232.204956,\n",
      "day 76: buy 1 unit at price 831.330017, total balance 8400.874939\n",
      "day 77, sell 1 unit at price 828.640015, investment 1.179520 %, total balance 9229.514954,\n",
      "day 78: buy 1 unit at price 829.280029, total balance 8400.234925\n",
      "day 79, sell 1 unit at price 823.210022, investment -0.976747 %, total balance 9223.444947,\n",
      "day 80: buy 1 unit at price 835.239990, total balance 8388.204957\n",
      "day 81, sell 1 unit at price 830.630005, investment 0.162789 %, total balance 9218.834962,\n",
      "day 83, sell 1 unit at price 827.780029, investment -0.893152 %, total balance 10046.614991,\n",
      "day 88: buy 1 unit at price 845.539978, total balance 9201.075013\n",
      "day 89, sell 1 unit at price 845.619995, investment 0.009463 %, total balance 10046.695008,\n",
      "day 91: buy 1 unit at price 848.780029, total balance 9197.914979\n",
      "day 92: buy 1 unit at price 852.119995, total balance 8345.794984\n",
      "day 93: buy 1 unit at price 848.400024, total balance 7497.394960\n",
      "day 96, sell 1 unit at price 817.580017, investment -3.675865 %, total balance 8314.974977,\n",
      "day 97: buy 1 unit at price 814.429993, total balance 7500.544984\n",
      "day 100, sell 1 unit at price 831.409973, investment -2.430411 %, total balance 8331.954957,\n",
      "day 101, sell 1 unit at price 831.500000, investment -1.991988 %, total balance 9163.454957,\n",
      "day 103, sell 1 unit at price 838.549988, investment 2.961580 %, total balance 10002.004945,\n",
      "day 104: buy 1 unit at price 834.570007, total balance 9167.434938\n",
      "day 105: buy 1 unit at price 831.409973, total balance 8336.024965\n",
      "day 106, sell 1 unit at price 827.880005, investment -0.801611 %, total balance 9163.904970,\n",
      "day 107: buy 1 unit at price 824.669983, total balance 8339.234987\n",
      "day 108: buy 1 unit at price 824.729980, total balance 7514.505007\n",
      "day 110: buy 1 unit at price 824.320007, total balance 6690.185000\n",
      "day 111: buy 1 unit at price 823.559998, total balance 5866.625002\n",
      "day 112, sell 1 unit at price 837.169983, investment 0.692800 %, total balance 6703.794985,\n",
      "day 113, sell 1 unit at price 836.820007, investment 1.473320 %, total balance 7540.614992,\n",
      "day 114, sell 1 unit at price 838.210022, investment 1.634479 %, total balance 8378.825014,\n",
      "day 115, sell 1 unit at price 841.650024, investment 2.102341 %, total balance 9220.475038,\n",
      "day 116, sell 1 unit at price 843.190002, investment 2.383555 %, total balance 10063.665040,\n",
      "day 122: buy 1 unit at price 912.570007, total balance 9151.095033\n",
      "day 123, sell 1 unit at price 916.440002, investment 0.424077 %, total balance 10067.535035,\n",
      "day 125: buy 1 unit at price 931.659973, total balance 9135.875062\n",
      "day 126, sell 1 unit at price 927.130005, investment -0.486225 %, total balance 10063.005067,\n",
      "day 144: buy 1 unit at price 966.950012, total balance 9096.055055\n",
      "day 145, sell 1 unit at price 975.599976, investment 0.894562 %, total balance 10071.655031,\n",
      "day 146: buy 1 unit at price 983.679993, total balance 9087.975038\n",
      "day 148: buy 1 unit at price 980.940002, total balance 8107.035036\n",
      "day 149, sell 1 unit at price 983.409973, investment -0.027450 %, total balance 9090.445009,\n",
      "day 151, sell 1 unit at price 942.900024, investment -3.877911 %, total balance 10033.345033,\n",
      "day 155: buy 1 unit at price 939.780029, total balance 9093.565004\n",
      "day 157, sell 1 unit at price 950.630005, investment 1.154523 %, total balance 10044.195009,\n",
      "day 160: buy 1 unit at price 965.590027, total balance 9078.604982\n",
      "day 161, sell 1 unit at price 952.270020, investment -1.379468 %, total balance 10030.875002,\n",
      "day 164: buy 1 unit at price 917.789978, total balance 9113.085024\n",
      "day 166, sell 1 unit at price 898.700012, investment -2.079993 %, total balance 10011.785036,\n",
      "day 168: buy 1 unit at price 906.690002, total balance 9105.095034\n",
      "day 171: buy 1 unit at price 930.090027, total balance 8175.005007\n",
      "day 172, sell 1 unit at price 943.830017, investment 4.096220 %, total balance 9118.835024,\n",
      "day 173, sell 1 unit at price 947.159973, investment 1.835300 %, total balance 10065.994997,\n",
      "day 176: buy 1 unit at price 965.400024, total balance 9100.594973\n",
      "day 179, sell 1 unit at price 972.919983, investment 0.778947 %, total balance 10073.514956,\n",
      "day 180: buy 1 unit at price 980.340027, total balance 9093.174929\n",
      "day 181, sell 1 unit at price 950.700012, investment -3.023442 %, total balance 10043.874941,\n",
      "day 194: buy 1 unit at price 914.390015, total balance 9129.484926\n",
      "day 195, sell 1 unit at price 922.669983, investment 0.905518 %, total balance 10052.154909,\n",
      "day 197: buy 1 unit at price 926.960022, total balance 9125.194887\n",
      "day 199: buy 1 unit at price 910.669983, total balance 8214.524904\n",
      "day 201, sell 1 unit at price 924.690002, investment -0.244889 %, total balance 9139.214906,\n",
      "day 202, sell 1 unit at price 927.000000, investment 1.793187 %, total balance 10066.214906,\n",
      "day 208: buy 1 unit at price 939.330017, total balance 9126.884889\n",
      "day 211, sell 1 unit at price 927.809998, investment -1.226408 %, total balance 10054.694887,\n",
      "day 214: buy 1 unit at price 929.080017, total balance 9125.614870\n",
      "day 216, sell 1 unit at price 935.090027, investment 0.646878 %, total balance 10060.704897,\n",
      "day 219: buy 1 unit at price 915.000000, total balance 9145.704897\n",
      "day 221: buy 1 unit at price 931.580017, total balance 8214.124880\n",
      "day 222, sell 1 unit at price 932.450012, investment 1.907105 %, total balance 9146.574892,\n",
      "day 223, sell 1 unit at price 928.530029, investment -0.327399 %, total balance 10075.104921,\n",
      "day 224: buy 1 unit at price 920.969971, total balance 9154.134950\n",
      "day 226: buy 1 unit at price 944.489990, total balance 8209.644960\n",
      "day 227: buy 1 unit at price 949.500000, total balance 7260.144960\n",
      "day 228, sell 1 unit at price 959.109985, investment 4.141287 %, total balance 8219.254945,\n",
      "day 229, sell 1 unit at price 953.270020, investment 0.929605 %, total balance 9172.524965,\n",
      "day 230: buy 1 unit at price 957.789978, total balance 8214.734987\n",
      "day 231: buy 1 unit at price 951.679993, total balance 7263.054994\n",
      "day 232, sell 1 unit at price 969.960022, investment 2.154821 %, total balance 8233.015016,\n",
      "day 233, sell 1 unit at price 978.890015, investment 2.202992 %, total balance 9211.905031,\n",
      "day 234, sell 1 unit at price 977.000000, investment 2.660559 %, total balance 10188.905031,\n"
     ]
    }
   ],
   "source": [
    "states_buy, states_sell, total_gains, invest = agent.buy(initial_money = initial_money)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize = (15,5))\n",
    "plt.plot(close, color='r', lw=2.)\n",
    "plt.plot(close, '^', markersize=10, color='m', label = 'buying signal', markevery = states_buy)\n",
    "plt.plot(close, 'v', markersize=10, color='k', label = 'selling signal', markevery = states_sell)\n",
    "plt.title('total gains %f, total investment %f%%'%(total_gains, invest))\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}


================================================
FILE: agent/11.double-duel-q-learning-agent.ipynb
================================================
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Date</th>\n",
       "      <th>Open</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Close</th>\n",
       "      <th>Adj Close</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2016-11-02</td>\n",
       "      <td>778.200012</td>\n",
       "      <td>781.650024</td>\n",
       "      <td>763.450012</td>\n",
       "      <td>768.700012</td>\n",
       "      <td>768.700012</td>\n",
       "      <td>1872400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2016-11-03</td>\n",
       "      <td>767.250000</td>\n",
       "      <td>769.950012</td>\n",
       "      <td>759.030029</td>\n",
       "      <td>762.130005</td>\n",
       "      <td>762.130005</td>\n",
       "      <td>1943200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016-11-04</td>\n",
       "      <td>750.659973</td>\n",
       "      <td>770.359985</td>\n",
       "      <td>750.560974</td>\n",
       "      <td>762.020020</td>\n",
       "      <td>762.020020</td>\n",
       "      <td>2134800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016-11-07</td>\n",
       "      <td>774.500000</td>\n",
       "      <td>785.190002</td>\n",
       "      <td>772.549988</td>\n",
       "      <td>782.520020</td>\n",
       "      <td>782.520020</td>\n",
       "      <td>1585100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016-11-08</td>\n",
       "      <td>783.400024</td>\n",
       "      <td>795.632996</td>\n",
       "      <td>780.190002</td>\n",
       "      <td>790.510010</td>\n",
       "      <td>790.510010</td>\n",
       "      <td>1350800</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Date        Open        High         Low       Close   Adj Close  \\\n",
       "0  2016-11-02  778.200012  781.650024  763.450012  768.700012  768.700012   \n",
       "1  2016-11-03  767.250000  769.950012  759.030029  762.130005  762.130005   \n",
       "2  2016-11-04  750.659973  770.359985  750.560974  762.020020  762.020020   \n",
       "3  2016-11-07  774.500000  785.190002  772.549988  782.520020  782.520020   \n",
       "4  2016-11-08  783.400024  795.632996  780.190002  790.510010  790.510010   \n",
       "\n",
       "    Volume  \n",
       "0  1872400  \n",
       "1  1943200  \n",
       "2  2134800  \n",
       "3  1585100  \n",
       "4  1350800  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('../dataset/GOOG-year.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import deque\n",
    "import random\n",
    "\n",
    "class Model:\n",
    "    def __init__(self, input_size, output_size, layer_size, learning_rate):\n",
    "        self.X = tf.placeholder(tf.float32, (None, input_size))\n",
    "        self.Y = tf.placeholder(tf.float32, (None, output_size))\n",
    "        feed = tf.layers.dense(self.X, layer_size, activation = tf.nn.relu)\n",
    "        tensor_action, tensor_validation = tf.split(feed,2,1)\n",
    "        feed_action = tf.layers.dense(tensor_action, output_size)\n",
    "        feed_validation = tf.layers.dense(tensor_validation, 1)\n",
    "        self.logits = feed_validation + tf.subtract(feed_action,tf.reduce_mean(feed_action,axis=1,keep_dims=True))\n",
    "        self.cost = tf.reduce_sum(tf.square(self.Y - self.logits))\n",
    "        self.optimizer = tf.train.AdamOptimizer(learning_rate = learning_rate).minimize(self.cost)\n",
    "        \n",
    "class Agent:\n",
    "\n",
    "    LEARNING_RATE = 0.003\n",
    "    BATCH_SIZE = 32\n",
    "    LAYER_SIZE = 500\n",
    "    OUTPUT_SIZE = 3\n",
    "    EPSILON = 0.5\n",
    "    DECAY_RATE = 0.005\n",
    "    MIN_EPSILON = 0.1\n",
    "    GAMMA = 0.99\n",
    "    MEMORIES = deque()\n",
    "    COPY = 1000\n",
    "    T_COPY = 0\n",
    "    MEMORY_SIZE = 300\n",
    "    \n",
    "    def __init__(self, state_size, window_size, trend, skip):\n",
    "        self.state_size = state_size\n",
    "        self.window_size = window_size\n",
    "        self.half_window = window_size // 2\n",
    "        self.trend = trend\n",
    "        self.skip = skip\n",
    "        tf.reset_default_graph()\n",
    "        self.model = Model(self.state_size, self.OUTPUT_SIZE, self.LAYER_SIZE, self.LEARNING_RATE)\n",
    "        self.model_negative = Model(self.state_size, self.OUTPUT_SIZE, self.LAYER_SIZE, self.LEARNING_RATE)\n",
    "        self.sess = tf.InteractiveSession()\n",
    "        self.sess.run(tf.global_variables_initializer())\n",
    "        self.trainable = tf.trainable_variables()\n",
    "    \n",
    "    def _assign(self):\n",
    "        for i in range(len(self.trainable)//2):\n",
    "            assign_op = self.trainable[i+len(self.trainable)//2].assign(self.trainable[i])\n",
    "            self.sess.run(assign_op)\n",
    "\n",
    "    def _memorize(self, state, action, reward, new_state, done):\n",
    "        self.MEMORIES.append((state, action, reward, new_state, done))\n",
    "        if len(self.MEMORIES) > self.MEMORY_SIZE:\n",
    "            self.MEMORIES.popleft()\n",
    "\n",
    "    def _select_action(self, state):\n",
    "        if np.random.rand() < self.EPSILON:\n",
    "            action = np.random.randint(self.OUTPUT_SIZE)\n",
    "        else:\n",
    "            action = self.get_predicted_action([state])\n",
    "        return action\n",
    "\n",
    "    def _construct_memories(self, replay):\n",
    "        states = np.array([a[0] for a in replay])\n",
    "        new_states = np.array([a[3] for a in replay])\n",
    "        Q = self.predict(states)\n",
    "        Q_new = self.predict(new_states)\n",
    "        Q_new_negative = self.sess.run(self.model_negative.logits, feed_dict={self.model_negative.X:new_states})\n",
    "        replay_size = len(replay)\n",
    "        X = np.empty((replay_size, self.state_size))\n",
    "        Y = np.empty((replay_size, self.OUTPUT_SIZE))\n",
    "        for i in range(replay_size):\n",
    "            state_r, action_r, reward_r, new_state_r, done_r = replay[i]\n",
    "            target = Q[i]\n",
    "            target[action_r] = reward_r\n",
    "            if not done_r:\n",
    "                target[action_r] += self.GAMMA * Q_new_negative[i, np.argmax(Q_new[i])]\n",
    "            X[i] = state_r\n",
    "            Y[i] = target\n",
    "        return X, Y\n",
    "\n",
    "    def predict(self, inputs):\n",
    "        return self.sess.run(self.model.logits, feed_dict={self.model.X:inputs})\n",
    "    \n",
    "    def get_predicted_action(self, sequence):\n",
    "        prediction = self.predict(np.array(sequence))[0]\n",
    "        return np.argmax(prediction)\n",
    "    \n",
    "    def get_state(self, t):\n",
    "        window_size = self.window_size + 1\n",
    "        d = t - window_size + 1\n",
    "        block = self.trend[d : t + 1] if d >= 0 else -d * [self.trend[0]] + self.trend[0 : t + 1]\n",
    "        res = []\n",
    "        for i in range(window_size - 1):\n",
    "            res.append(block[i + 1] - block[i])\n",
    "        return np.array(res)\n",
    "    \n",
    "    def buy(self, initial_money):\n",
    "        starting_money = initial_money\n",
    "        states_sell = []\n",
    "        states_buy = []\n",
    "        inventory = []\n",
    "        state = self.get_state(0)\n",
    "        for t in range(0, len(self.trend) - 1, self.skip):\n",
    "            action = self._select_action(state)\n",
    "            next_state = self.get_state(t + 1)\n",
    "            \n",
    "            if action == 1 and initial_money >= self.trend[t]:\n",
    "                inventory.append(self.trend[t])\n",
    "                initial_money -= self.trend[t]\n",
    "                states_buy.append(t)\n",
    "                print('day %d: buy 1 unit at price %f, total balance %f'% (t, self.trend[t], initial_money))\n",
    "            \n",
    "            elif action == 2 and len(inventory):\n",
    "                bought_price = inventory.pop(0)\n",
    "                initial_money += self.trend[t]\n",
    "                states_sell.append(t)\n",
    "                try:\n",
    "                    invest = ((close[t] - bought_price) / bought_price) * 100\n",
    "                except:\n",
    "                    invest = 0\n",
    "                print(\n",
    "                    'day %d, sell 1 unit at price %f, investment %f %%, total balance %f,'\n",
    "                    % (t, close[t], invest, initial_money)\n",
    "                )\n",
    "            \n",
    "            state = next_state\n",
    "        invest = ((initial_money - starting_money) / starting_money) * 100\n",
    "        total_gains = initial_money - starting_money\n",
    "        return states_buy, states_sell, total_gains, invest\n",
    "            \n",
    "    \n",
    "    def train(self, iterations, checkpoint, initial_money):\n",
    "        for i in range(iterations):\n",
    "            total_profit = 0\n",
    "            inventory = []\n",
    "            state = self.get_state(0)\n",
    "            starting_money = initial_money\n",
    "            for t in range(0, len(self.trend) - 1, self.skip):\n",
    "                if (self.T_COPY + 1) % self.COPY == 0:\n",
    "                    self._assign()\n",
    "                \n",
    "                action = self._select_action(state)\n",
    "                next_state = self.get_state(t + 1)\n",
    "                \n",
    "                if action == 1 and starting_money >= self.trend[t]:\n",
    "                    inventory.append(self.trend[t])\n",
    "                    starting_money -= self.trend[t]\n",
    "                \n",
    "                elif action == 2 and len(inventory) > 0:\n",
    "                    bought_price = inventory.pop(0)\n",
    "                    total_profit += self.trend[t] - bought_price\n",
    "                    starting_money += self.trend[t]\n",
    "                    \n",
    "                invest = ((starting_money - initial_money) / initial_money)\n",
    "                \n",
    "                self._memorize(state, action, invest, next_state, starting_money < initial_money)\n",
    "                batch_size = min(len(self.MEMORIES), self.BATCH_SIZE)\n",
    "                state = next_state\n",
    "                replay = random.sample(self.MEMORIES, batch_size)\n",
    "                X, Y = self._construct_memories(replay)\n",
    "                \n",
    "                cost, _ = self.sess.run([self.model.cost, self.model.optimizer], \n",
    "                                        feed_dict={self.model.X: X, self.model.Y:Y})\n",
    "                self.T_COPY += 1\n",
    "                self.EPSILON = self.MIN_EPSILON + (1.0 - self.MIN_EPSILON) * np.exp(-self.DECAY_RATE * i)\n",
    "            if (i+1) % checkpoint == 0:\n",
    "                print('epoch: %d, total rewards: %f.3, cost: %f, total money: %f'%(i + 1, total_profit, cost,\n",
    "                                                                                  starting_money))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From <ipython-input-3-42f2d1e26a9d>:12: calling reduce_mean (from tensorflow.python.ops.math_ops) with keep_dims is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "keep_dims is deprecated, use keepdims instead\n",
      "epoch: 10, total rewards: 1486.684997.3, cost: 0.694152, total money: 10514.124999\n",
      "epoch: 20, total rewards: 313.279660.3, cost: 0.878157, total money: 8354.909665\n",
      "epoch: 30, total rewards: 752.595089.3, cost: 0.320037, total money: 10752.595089\n",
      "epoch: 40, total rewards: 1159.299987.3, cost: 0.318166, total money: 10186.739989\n",
      "epoch: 50, total rewards: 993.220279.3, cost: 0.391151, total money: 4149.310245\n",
      "epoch: 60, total rewards: 1616.499880.3, cost: 0.307440, total money: 9630.939883\n",
      "epoch: 70, total rewards: 941.484560.3, cost: 0.332979, total money: 6969.054506\n",
      "epoch: 80, total rewards: 904.899903.3, cost: 0.718111, total money: 1132.559876\n",
      "epoch: 90, total rewards: 346.619873.3, cost: 0.482044, total money: 542.599852\n",
      "epoch: 100, total rewards: 141.554626.3, cost: 0.238426, total money: 6115.974608\n",
      "epoch: 110, total rewards: -159.529845.3, cost: 0.202412, total money: 8852.270143\n",
      "epoch: 120, total rewards: -37.579779.3, cost: 0.433529, total money: 8945.780206\n",
      "epoch: 130, total rewards: 1049.544800.3, cost: 0.408910, total money: 8099.664795\n",
      "epoch: 140, total rewards: 59.114809.3, cost: 0.028664, total money: 7098.904848\n",
      "epoch: 150, total rewards: 96.424866.3, cost: 0.070552, total money: 9079.784851\n",
      "epoch: 160, total rewards: 74.179754.3, cost: 0.044092, total money: 10074.179754\n",
      "epoch: 170, total rewards: 80.999883.3, cost: 0.018813, total money: 8047.249883\n",
      "epoch: 180, total rewards: 62.700011.3, cost: 0.083292, total money: 10062.700011\n",
      "epoch: 190, total rewards: 70.424991.3, cost: 0.013884, total money: 9053.315006\n",
      "epoch: 200, total rewards: 10.620115.3, cost: 0.030838, total money: 10010.620115\n"
     ]
    }
   ],
   "source": [
    "close = df.Close.values.tolist()\n",
    "initial_money = 10000\n",
    "window_size = 30\n",
    "skip = 1\n",
    "batch_size = 32\n",
    "agent = Agent(state_size = window_size, \n",
    "              window_size = window_size, \n",
    "              trend = close, \n",
    "              skip = skip)\n",
    "agent.train(iterations = 200, checkpoint = 10, initial_money = initial_money)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "day 1: buy 1 unit at price 762.130005, total balance 9237.869995\n",
      "day 2, sell 1 unit at price 762.020020, investment -0.014431 %, total balance 9999.890015,\n",
      "day 11: buy 1 unit at price 771.229980, total balance 9228.660035\n",
      "day 12: buy 1 unit at price 760.539978, total balance 8468.120057\n",
      "day 13, sell 1 unit at price 769.200012, investment -0.263212 %, total balance 9237.320069,\n",
      "day 15, sell 1 unit at price 760.989990, investment 0.059170 %, total balance 9998.310059,\n",
      "day 34: buy 1 unit at price 794.559998, total balance 9203.750061\n",
      "day 35, sell 1 unit at price 791.260010, investment -0.415323 %, total balance 9995.010071,\n",
      "day 36: buy 1 unit at price 789.909973, total balance 9205.100098\n",
      "day 37, sell 1 unit at price 791.549988, investment 0.207620 %, total balance 9996.650086,\n",
      "day 38: buy 1 unit at price 785.049988, total balance 9211.600098\n",
      "day 40, sell 1 unit at price 771.820007, investment -1.685241 %, total balance 9983.420105,\n",
      "day 54: buy 1 unit at price 819.309998, total balance 9164.110107\n",
      "day 55, sell 1 unit at price 823.869995, investment 0.556566 %, total balance 9987.980102,\n",
      "day 62: buy 1 unit at price 798.530029, total balance 9189.450073\n",
      "day 64, sell 1 unit at price 801.340027, investment 0.351896 %, total balance 9990.790100,\n",
      "day 68: buy 1 unit at price 813.669983, total balance 9177.120117\n",
      "day 69, sell 1 unit at price 819.239990, investment 0.684554 %, total balance 9996.360107,\n",
      "day 72: buy 1 unit at price 824.159973, total balance 9172.200134\n",
      "day 73, sell 1 unit at price 828.070007, investment 0.474427 %, total balance 10000.270141,\n",
      "day 74: buy 1 unit at price 831.659973, total balance 9168.610168\n",
      "day 75, sell 1 unit at price 830.760010, investment -0.108213 %, total balance 9999.370178,\n",
      "day 79: buy 1 unit at price 823.210022, total balance 9176.160156\n",
      "day 80, sell 1 unit at price 835.239990, investment 1.461349 %, total balance 10011.400146,\n",
      "day 90: buy 1 unit at price 847.200012, total balance 9164.200134\n",
      "day 91, sell 1 unit at price 848.780029, investment 0.186499 %, total balance 10012.980163,\n",
      "day 93: buy 1 unit at price 848.400024, total balance 9164.580139\n",
      "day 94: buy 1 unit at price 830.460022, total balance 8334.120117\n",
      "day 95, sell 1 unit at price 829.590027, investment -2.217114 %, total balance 9163.710144,\n",
      "day 96, sell 1 unit at price 817.580017, investment -1.550948 %, total balance 9981.290161,\n",
      "day 100: buy 1 unit at price 831.409973, total balance 9149.880188\n",
      "day 101, sell 1 unit at price 831.500000, investment 0.010828 %, total balance 9981.380188,\n",
      "day 104: buy 1 unit at price 834.570007, total balance 9146.810181\n",
      "day 106: buy 1 unit at price 827.880005, total balance 8318.930176\n",
      "day 107, sell 1 unit at price 824.669983, investment -1.186242 %, total balance 9143.600159,\n",
      "day 108, sell 1 unit at price 824.729980, investment -0.380493 %, total balance 9968.330139,\n",
      "day 110: buy 1 unit at price 824.320007, total balance 9144.010132\n",
      "day 111, sell 1 unit at price 823.559998, investment -0.092198 %, total balance 9967.570130,\n",
      "day 115: buy 1 unit at price 841.650024, total balance 9125.920106\n",
      "day 116, sell 1 unit at price 843.190002, investment 0.182971 %, total balance 9969.110108,\n",
      "day 125: buy 1 unit at price 931.659973, total balance 9037.450135\n",
      "day 126, sell 1 unit at price 927.130005, investment -0.486225 %, total balance 9964.580140,\n",
      "day 127: buy 1 unit at price 934.299988, total balance 9030.280152\n",
      "day 128, sell 1 unit at price 932.169983, investment -0.227979 %, total balance 9962.450135,\n",
      "day 141: buy 1 unit at price 971.469971, total balance 8990.980164\n",
      "day 142, sell 1 unit at price 975.880005, investment 0.453955 %, total balance 9966.860169,\n",
      "day 152: buy 1 unit at price 953.400024, total balance 9013.460145\n",
      "day 153, sell 1 unit at price 950.760010, investment -0.276905 %, total balance 9964.220155,\n",
      "day 156: buy 1 unit at price 957.369995, total balance 9006.850160\n",
      "day 157, sell 1 unit at price 950.630005, investment -0.704011 %, total balance 9957.480165,\n",
      "day 166: buy 1 unit at price 898.700012, total balance 9058.780153\n",
      "day 167, sell 1 unit at price 911.710022, investment 1.447648 %, total balance 9970.490175,\n",
      "day 172: buy 1 unit at price 943.830017, total balance 9026.660158\n",
      "day 173, sell 1 unit at price 947.159973, investment 0.352813 %, total balance 9973.820131,\n",
      "day 185: buy 1 unit at price 930.500000, total balance 9043.320131\n",
      "day 186: buy 1 unit at price 930.830017, total balance 8112.490114\n",
      "day 187, sell 1 unit at price 930.390015, investment -0.011820 %, total balance 9042.880129,\n",
      "day 188, sell 1 unit at price 923.650024, investment -0.771354 %, total balance 9966.530153,\n",
      "day 193: buy 1 unit at price 907.239990, total balance 9059.290163\n",
      "day 194, sell 1 unit at price 914.390015, investment 0.788107 %, total balance 9973.680178,\n",
      "day 197: buy 1 unit at price 926.960022, total balance 9046.720156\n",
      "day 199, sell 1 unit at price 910.669983, investment -1.757362 %, total balance 9957.390139,\n",
      "day 211: buy 1 unit at price 927.809998, total balance 9029.580141\n",
      "day 212, sell 1 unit at price 935.950012, investment 0.877336 %, total balance 9965.530153,\n",
      "day 213: buy 1 unit at price 926.500000, total balance 9039.030153\n",
      "day 214, sell 1 unit at price 929.080017, investment 0.278469 %, total balance 9968.110170,\n"
     ]
    }
   ],
   "source": [
    "states_buy, states_sell, total_gains, invest = agent.buy(initial_money = initial_money)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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