Explore Stock Analysis

Tiny sandbox for experimenting with stock-price forecasting using engineered technical features and a basic PyTorch MLP.

• Data: CSVs in data/ (e.g., nasdaq_19900101_20230630.csv) with OHLCV columns and Date index.
• Features: Rolling means/std devs, volume ratios, and lagged returns built via prepare_data_features in helpers.py.
• Model: NasdaqBasicMLPRunner in nasdaq_basic_mlp.py trains a 2-layer MLP on standardized features and reports MSE/MAE/R².

Quickstart (uv)

uv sync  # creates .venv and installs deps from uv.lock/pyproject.toml
uv run nasdaq_basic_mlp.py  # trains and prints test metrics

Run on a remote GPU host

• Ensure rsync is installed: apt update && apt install rsync -y
• Install uv: curl -LsSf https://astral.sh/uv/install.sh | sh
• Sync project to remote host: bin/rspec 1.1.1.1:22 /workspace (/workspace is implied/default)
• SSH into remote host and run: cd /workspace && uv sync && uv run python nasdaq_basic_mlp.py

Docker (GPU)

• Build: docker build -t explore-stock-analysis:latest .
• Run (NVIDIA): docker run --rm --gpus all explore-stock-analysis:latest

Notes

• Adjust train/test splits, learning rate, or epochs in the NasdaqBasicMLPRunner constructor.
• Swap data/*.csv or add new ones; filenames map to load_stock_data("<name>") without the .csv suffix.