Skip to content

Latest commit

 

History

History
29 lines (18 loc) · 1009 Bytes

README.md

File metadata and controls

29 lines (18 loc) · 1009 Bytes

Instructions for running

To install requirements call
    > pip install -r requirements.txt

Just a warning, all of these can be very slow to run (especially without a GPU).

To train:

> python3 data_parse.py

The -s option skips the creation of the data file if included (need to run at least once without -s initially)
the -n option says to skip training
the -g option is for when training is going to happen on a gpu
the -m option specifies a model to load and continue training if one is availible

To forecast/evaluate:

> python3 machine_learn.py -m path/to/model

the -m option specifies which model to evaluate
can include -n option to skip forecasting and just graph the predictions from the neural net

This script will generate a plot at the end that shows how well it did in predicting and forecasting for the set given. the plot is interactive and one can zoom in on any areas of interest.

To evaluate battery sizes:

> python3 battery.py