More to come
Install requirements (obvs)
pip install -r requirements.txt
Default settings will direct data collection to data directory in project root. To specify alternate directory, change DATA_DIR in settings.py.
Collect and store data from pandas datareader
python utils/collect_from_pandas
To get most recent set of economic data from multipl.com:
cd multipl
scrapy crawl multipl
This will store a collection of json objects in your data folder.
The machine learnign model currently relies on the SMOTE algorithm to upsample rare data classes and create a more balanced traiing set.
I used an implementation from the "UnbalancedDataset" library but because the package is not available in PyPi, I simply included the library in the utils folder.
The source for this library can be found here: https://github.com/fmfn/UnbalancedDataset
More to come... duh
Hi Dragon