kyat prediction with sklearn
scrape.py
This script scrapes the web for data concerning currency trading prices of the Myanmar Kyat. Output is rates.pkl, a Pandas DataFrame with all the information queried previously.
scrape.py reads the rates.pkl and for every queried date that hasn't already been recorded, it appends a Series with the rates for that day. It then sorts the DataFrame by the date, and outputs a Pickle.
train.py
This script runs linear and ridge regression on the data to produce a prediction for tomorrow.
.pkl files
These are python pickle objects. I am using Python 3, so these are generated with the library dill. db.pkl is a pandas DataFrame object. index.pkl is a list of dates that correspond to the rows in db.pkl. The columns are given in currencies.pkl.
maps.ipynb
This Jupyter Notebook is just for making some images to put on the website.