Quick Start

Load your first materials science dataset in under 5 minutes.

1. Install

pip install foundry-ml

2. Connect

from foundry import Foundry

f = Foundry()

For cloud environments (Colab, remote Jupyter):

f = Foundry(no_browser=True, no_local_server=True)

3. Search

Find datasets by keyword:

results = f.search("band gap", limit=5)
results

Output:

                                    dataset_name  ...
0                   foundry_oqmd_band_gaps_v1.1  ...
1                   foundry_aflow_band_gaps_v1.1  ...
2            foundry_experimental_band_gaps_v1.1  ...
...

4. Load

Get a dataset and load its data:

# Get the first result
dataset = results.iloc[0].FoundryDataset

# Load training data
data = dataset.get_as_dict()
X_train, y_train = data['train']

print(f"Samples: {len(X_train)}")

5. Understand

Check what fields the dataset contains:

schema = dataset.get_schema()

print(f"Dataset: {schema['name']}")
print(f"Fields:")
for field in schema['fields']:
    print(f"  - {field['name']} ({field['role']})")

6. Cite

Get the citation for publications:

print(dataset.get_citation())

Complete Example

from foundry import Foundry

# Connect
f = Foundry()

# Search
results = f.search("band gap", limit=5)

# Load
dataset = results.iloc[0].FoundryDataset
X, y = dataset.get_as_dict()['train']

# Use with sklearn
from sklearn.ensemble import RandomForestRegressor
model = RandomForestRegressor()
# model.fit(X, y)  # Train your model

# Cite
print(dataset.get_citation())