Skip to content

Getting Started

Requirements: python 3.10+

Install

# basic install
pip install stocktracer

# with tensorflow dependencies for analysis modules
pip install stocktracer[tensorflow]

# Perform analysis
stocktracer analyze --tickers aapl,msft > report.txt

# Help
stocktracer

# Clone Repository
git clone https://github.com/gyund/fundamental-analysis.git
cd fundamental-analysis.git

# Normal development install
poetry install --sync # (1)!

# With mkdocs
poetry install --sync --with docs

# with mkdocs and tensorflow dependency support
poetry install --sync --extras "tensorflow"  --with docs

# Perform analysis
PYTHONPATH=src poetry run python -m stocktracer analyze --tickers aapl,msft
PYTHONPATH=src poetry run python -m stocktracer analyze --tickers aapl,msft -a stocktracer.analysis.diluted_eps
PYTHONPATH=src poetry run python -m stocktracer analyze --tickers aapl,msft -a stocktracer.analysis.diluted_eps --report-format csv
PYTHONPATH=src poetry run python -m stocktracer analyze --tickers aapl,msft -a stocktracer.analysis.diluted_eps --report-format json --report-file my_results.json

# Help
PYTHONPATH=src poetry run python -m stocktracer

# Run Unit Tests
poetry run pytest
  1. Make sure you have poetry installed.

Generating a Basic Report

# Perform analysis
stocktracer analyze --tickers appl,msft > report.txt

# Perform analysis
python -m stocktracer analyze --tickers aapl,msft > report.txt

Tip

If you want to figure out a list of tags you can filter the reports on, run the default analysis report. This shows the annual report and will then filter out any columns that contain null or NaN values. From here, you can establish what algorithms you can use and apply consistently across the stocks of interest. You may find that different sectors or 10-K/10-Q reports will have different data sets.

Plugins

If you wish to use your own analysis plugin, create your own module that implements this interface:

from ticker.data.sec import DataSelector as SecDataSelector
from ticker.filter import Selectors,SecFilter

def analyze(options: Options) -> None:
    print("This is where we would start to process information, but we're not right now")

Then call the tool in the following manner:

python -m ticker analyze --tickers aapl,msft --analysis_plugin 'mypkg.analysis'

Testing

Run tests by running the following:

pytest

If you wish to run tests using real network resources, such as downloading real reports and processing them, run the following:

pytest --run-webtest

Note that all data sets will be cached in the directory ${cwd}/.ticker-cache/ for both real and test runs. Expiry for quarterly reports are cached for 5 years and ticker mappings for CIK -> Ticker conversion are cached on a yearly basis. You generally won't be researching companies with less than a year's worth of reports though this could cause recently listed companies to lack CIK -> Ticker conversions for up to two years from poor timing. Just delete ${cwd}/.ticker-cache/tickers.sqlite to get the latest.

Last update: June 3, 2023