This is a project to predict price movement of a stock based on it's candle stick chart. I have some example candle stick charts which denotes if the price should go up or down. The idea is to find the closest match for the sample candle stick chart images to predict what will happen. The image embeddings are generated by openai api, closest match is calculated using sklearn's cosine similarity and numpy.
- Assuming you have Python 3+ installed
- Create a
.envfile in the root - Add your openai api key under
OPENAI_API_KEY - Run
pip3 install -r src/requirements.txtto install the python packages - Run
python3 src/main.pyUpon running the command, follow the terminal instructions.
Note that the candle charts can be only seen on MacOS iTerm2.