data/README.md

# Data Directory

This folder contains the training and test datasets for the Milk Spoilage Classification model.

## Files

- `train_df.csv` - Training dataset
- `test_df.csv` - Test dataset
- `sensory_dat.csv` - Full sensory analysis data
- `sensory_train_dat.csv` - Training sensory data
- `sensory_test_dat.csv` - Test sensory data

## Data Format

The datasets contain microbial count measurements:

- `SPC_D7`, `SPC_D14`, `SPC_D21` - Standard Plate Count at days 7, 14, and 21 (log CFU/mL)
- `TGN_D7`, `TGN_D14`, `TGN_D21` - Total Gram-Negative count at days 7, 14, and 21 (log CFU/mL)
- `spoilagetype` - Target variable (PPC, no spoilage, spore spoilage)

## Usage

These files are used by `scripts/prepare_model.py` to train the classification model.