Right now a counter is limiting the number of tickers in the historical data folder being used to train and test the MLP. This needs to be updated so that more of the tickers are used. Ideally all ~7000 of them.
This should be done in the dataset.formatData() method, iterating through the stockFiles generator to read through the .csv files, ignoring empty ones, and uniformly formatting everything so that tensorFlow is happy.
Right now a counter is limiting the number of tickers in the historical data folder being used to train and test the MLP. This needs to be updated so that more of the tickers are used. Ideally all ~7000 of them.
This should be done in the dataset.formatData() method, iterating through the stockFiles generator to read through the .csv files, ignoring empty ones, and uniformly formatting everything so that tensorFlow is happy.