# Example: HAI 21.03 Feature Split

This folder contains a small, reproducible example that inspects the HAI 21.03
CSV (train1) and produces a continuous/discrete split using a simple heuristic.

## Files
- analyze_hai21_03.py: reads a sample of the data and writes results.
- data_utils.py: CSV loading, vocab, normalization, and batching helpers.
- feature_split.json: column split for HAI 21.03.
- hybrid_diffusion.py: hybrid model + diffusion utilities.
- prepare_data.py: compute vocab and normalization stats.
- train_stub.py: end-to-end scaffold for loss computation.
- train.py: minimal training loop with checkpoints.
- sample.py: minimal sampling loop.
- model_design.md: step-by-step design notes.
- results/feature_split.txt: comma-separated feature lists.
- results/summary.txt: basic stats (rows sampled, column counts).

## Run
```
python /home/anay/Dev/diffusion/mask-ddpm/example/analyze_hai21_03.py
```

Prepare vocab + stats (writes to `example/results`):
```
python /home/anay/Dev/diffusion/mask-ddpm/example/prepare_data.py
```

Train a small run:
```
python /home/anay/Dev/diffusion/mask-ddpm/example/train.py
```

Sample from the trained model:
```
python /home/anay/Dev/diffusion/mask-ddpm/example/sample.py
```

## Notes
- Heuristic: integer-like values with low cardinality (<=10) are treated as
  discrete. All other numeric columns are continuous.
- The script only samples the first 5000 rows to stay fast.
- `prepare_data.py` runs without PyTorch, but `train.py` and `sample.py` require it.
- `train.py` and `sample.py` auto-select GPU if available; otherwise they fall back to CPU.