Software

For stock price prediction, our software can be simply used with the following lines of codes:

Step 1: Clone this repo with:

git clone -b main https://github.com/CS205-Project-S21/Parallel-LSTM.git

Step 2: Install related packages. We recommend the user to first create a virtual environment with conda and do this step in the virtual environment:

conda create -n stock_price python==3.8
conda activate stock_price

Then run:

pip install Pararrel-LSTM/requirements.txt

Step 3: Change into working directory and run command:

cd Pararrel-LSTM/utils
python price_predictor.py --ticker <stock-ticker>

The <stock-ticker> here can be selected from one in the following list: [BTC-USD, MARA, RIOT, COG, DVN, HFC].

Then the software will present the prediction in great plots. In the left figure, the price on the left side of the dash line represents the true price in the past 10 days, on the right side of the dash line represents the predicted price in the coming 5 days. The figure on the right side represents the sentiment score from news calculated in the past 10 days, which will be helpful to the user to understand the trend on the left.