# Akademy
Akademy is a module containing composable object classes for developing
reinforcement learning algorithms focused on quantitative trading and
time-series forecasting. This module is a work-in-progress and should, at no
time, be assumed to be designed well or be free of bugs.
# Overview
Akademy is designed using an `Agent`-`Environment` model such that `Agent`-class
objects ingest information from `Environment`-class objects (`Env`), produce
an `Action`, which is then applied to the `Environment` which results in a
change in `State` and possible reward to offer feedback to the agent.
*Note*: this module does not provide any training routines -- only the object class
that can be used to support the implementation of custom training routines.
# Getting Started
To install `akademy` use the following command in the desired Python 3.7+
environment:
`pip install akademy`
Once installed, developers will have access to `Agent`, `TradeEnv`, and `Network`
class objects in which to design Reinforcement Learning algorithms to train models.
Sample training routine:
```python
from akademy.models.envs import TradeEnv
from akademy.models.agents import DQNAgent
from akademy.common.utils import load_spy_daily
# loads the dataset used during training
data = load_spy_daily(count=2500)
# load the Trading Environment
env = TradeEnv(
data=data,
window=50,
asset="spy",
)
# load the agent to train
agent = DQNAgent(
action_count=env.action_space.n,
state_shape=env.observation_space.shape
)
# load user-defined training routine
training_routine(
agent=agent,
env=env
)
```
## Tests
Unit testing can be run via the following command:
`python -m unittest`
For detailed information the `--verbose` flag can be used. For more detailed
usage consult the `unittest` module documentation.
## Available Data
This module comes with minimal data for Agents and Environments to train on.
The current data available is listed below, along with sources for the most
up-to-date versions as well:
### 1. S&P500
Location: `/data/SPY.CSV`\
Start: `1993-01-29`\
End: `2023-01-23`\
Total Rows: `7,454` (excludes header)\
Header: `Date,Open,High,Low,Close,Adj Close,Volume`\
Source: https://finance.yahoo.com/quote/SPY/history?p=SPY
*note*: Any data can be used easily enough via conversion into a Pandas DataFrame
object, but must contain information for `date` and pricing data for
`open`, `high`, `low`, and `close` as well as `volume` such that each row has
at least those 6 features or the latter 5 and an index representative of date.
# Notes
## Gym vs. Gymnasium
The `Gym` project by OpenAI has been sunset and now maintained as `Gymnasium`
by the [Farama-Foundation](https://github.com/Farama-Foundation/Gymnasium). The
`Env` classes present here make use of the newer `Gymnasium` package which, among
other differences, produces an extra item in the `step` method indicating whether
an environment has been truncated. [See here](https://github.com/Farama-Foundation/Gymnasium/blob/main/gymnasium/core.py#L63)
## PyTorch
PyTorch requires some additional consideration for setup depending on use-case.
Akademy uses an approach whereby CPU-based training and inferences are possible
via parameterized function calls. However, GPU use (e.g. CUDA) requires local
considerations. [See here] (https://pytorch.org/get-started/locally/) for a more
in-depth discussion and guide.
This module currently uses the 1.* version, though a 2.* version release
is imminent and an upgrade to that version is planned.
Raw data
{
"_id": null,
"home_page": "https://github.com/alphazwest/akademy",
"name": "akademy",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": "",
"keywords": "reinforcement learning,quantitative trading,fintech,trading bot,algorithmic trading,finance,automated trading,neural networks,artificial intelligence,machine learning",
"author": "Zack West",
"author_email": "Zack West <alphazwest@gmail.com>",
"download_url": "https://files.pythonhosted.org/packages/29/5c/3f9a9bdbb9e0cf44d61fd79f3700626257893cd587abc9b4c3ad9db2875e/akademy-0.1.51.tar.gz",
"platform": null,
"description": "# Akademy \r\n\r\nAkademy is a module containing composable object classes for developing \r\nreinforcement learning algorithms focused on quantitative trading and \r\ntime-series forecasting. This module is a work-in-progress and should, at no\r\ntime, be assumed to be designed well or be free of bugs.\r\n\r\n# Overview\r\nAkademy is designed using an `Agent`-`Environment` model such that `Agent`-class\r\nobjects ingest information from `Environment`-class objects (`Env`), produce\r\nan `Action`, which is then applied to the `Environment` which results in a\r\nchange in `State` and possible reward to offer feedback to the agent.\r\n\r\n*Note*: this module does not provide any training routines -- only the object class\r\nthat can be used to support the implementation of custom training routines.\r\n\r\n# Getting Started\r\n\r\nTo install `akademy` use the following command in the desired Python 3.7+\r\nenvironment:\r\n\r\n`pip install akademy`\r\n\r\nOnce installed, developers will have access to `Agent`, `TradeEnv`, and `Network`\r\nclass objects in which to design Reinforcement Learning algorithms to train models.\r\n\r\nSample training routine:\r\n\r\n```python\r\nfrom akademy.models.envs import TradeEnv\r\nfrom akademy.models.agents import DQNAgent\r\nfrom akademy.common.utils import load_spy_daily\r\n\r\n# loads the dataset used during training\r\ndata = load_spy_daily(count=2500)\r\n\r\n# load the Trading Environment\r\nenv = TradeEnv(\r\n data=data,\r\n window=50,\r\n asset=\"spy\",\r\n)\r\n\r\n# load the agent to train\r\nagent = DQNAgent(\r\n action_count=env.action_space.n,\r\n state_shape=env.observation_space.shape\r\n)\r\n\r\n# load user-defined training routine\r\ntraining_routine(\r\n agent=agent,\r\n env=env\r\n)\r\n```\r\n\r\n## Tests\r\nUnit testing can be run via the following command:\r\n\r\n`python -m unittest`\r\n\r\nFor detailed information the `--verbose` flag can be used. For more detailed \r\nusage consult the `unittest` module documentation.\r\n\r\n## Available Data\r\nThis module comes with minimal data for Agents and Environments to train on.\r\nThe current data available is listed below, along with sources for the most\r\nup-to-date versions as well:\r\n\r\n### 1. S&P500 \r\nLocation: `/data/SPY.CSV`\\\r\nStart: `1993-01-29`\\\r\nEnd: `2023-01-23`\\\r\nTotal Rows: `7,454` (excludes header)\\\r\nHeader: `Date,Open,High,Low,Close,Adj Close,Volume`\\\r\nSource: https://finance.yahoo.com/quote/SPY/history?p=SPY\r\n\r\n*note*: Any data can be used easily enough via conversion into a Pandas DataFrame\r\nobject, but must contain information for `date` and pricing data for\r\n`open`, `high`, `low`, and `close` as well as `volume` such that each row has\r\nat least those 6 features or the latter 5 and an index representative of date.\r\n\r\n# Notes\r\n\r\n## Gym vs. Gymnasium\r\nThe `Gym` project by OpenAI has been sunset and now maintained as `Gymnasium` \r\nby the [Farama-Foundation](https://github.com/Farama-Foundation/Gymnasium). The\r\n`Env` classes present here make use of the newer `Gymnasium` package which, among\r\nother differences, produces an extra item in the `step` method indicating whether\r\nan environment has been truncated. [See here](https://github.com/Farama-Foundation/Gymnasium/blob/main/gymnasium/core.py#L63)\r\n\r\n## PyTorch\r\nPyTorch requires some additional consideration for setup depending on use-case.\r\nAkademy uses an approach whereby CPU-based training and inferences are possible\r\nvia parameterized function calls. However, GPU use (e.g. CUDA) requires local\r\nconsiderations. [See here] (https://pytorch.org/get-started/locally/) for a more\r\nin-depth discussion and guide.\r\n\r\nThis module currently uses the 1.* version, though a 2.* version release\r\nis imminent and an upgrade to that version is planned.\r\n",
"bugtrack_url": null,
"license": "BSD-3-Clause",
"summary": "akademy: A Reinforcement Learning Framework",
"version": "0.1.51",
"split_keywords": [
"reinforcement learning",
"quantitative trading",
"fintech",
"trading bot",
"algorithmic trading",
"finance",
"automated trading",
"neural networks",
"artificial intelligence",
"machine learning"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "aedab560d3a55d2d13243dbf7c6886d5997936691c36a7251010e33f14b9825f",
"md5": "8b1d5b0750669da8cddd2fb3053ac820",
"sha256": "8ca049c34c33f87ee88a1a0b564e18004f885076d8436bd6d987f877279e1118"
},
"downloads": -1,
"filename": "akademy-0.1.51-py3-none-any.whl",
"has_sig": false,
"md5_digest": "8b1d5b0750669da8cddd2fb3053ac820",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.7",
"size": 27318,
"upload_time": "2023-03-29T00:16:04",
"upload_time_iso_8601": "2023-03-29T00:16:04.088660Z",
"url": "https://files.pythonhosted.org/packages/ae/da/b560d3a55d2d13243dbf7c6886d5997936691c36a7251010e33f14b9825f/akademy-0.1.51-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "295c3f9a9bdbb9e0cf44d61fd79f3700626257893cd587abc9b4c3ad9db2875e",
"md5": "0b764f13693fafbf056b544e26a49e00",
"sha256": "fe58357f4869982fc3116cdccee885eadd9cd8e1e60fd32b652a1a2d150a1a74"
},
"downloads": -1,
"filename": "akademy-0.1.51.tar.gz",
"has_sig": false,
"md5_digest": "0b764f13693fafbf056b544e26a49e00",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 23601,
"upload_time": "2023-03-29T00:16:07",
"upload_time_iso_8601": "2023-03-29T00:16:07.247970Z",
"url": "https://files.pythonhosted.org/packages/29/5c/3f9a9bdbb9e0cf44d61fd79f3700626257893cd587abc9b4c3ad9db2875e/akademy-0.1.51.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-03-29 00:16:07",
"github": true,
"gitlab": false,
"bitbucket": false,
"github_user": "alphazwest",
"github_project": "akademy",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [],
"lcname": "akademy"
}
JSON
