Installation#

To ensure that CRISP functions optimally, you must have the Python Environment(≥3.11) with the following software versions or higher installed on your system:

  • ase: 3.19.0

  • sklearn: 1.4.2

  • seaborn: 0.12.2

  • joblib: 1.2.0

  • statsmodels: 0.14.0

  • pandas: 2.0.3

  • plotly: 5.9.0

  • networkx: 3.1

  • fpsample>=0.3.3

  • dscribe>=2.0.0

You can install all the required packages using the following command:

pip install ase>=3.23.0 dscribe>=2.0.0 scikit-learn>=1.4.2 seaborn>=0.12.2 joblib>=1.2.0 fpsample>=0.3.3 statsmodels>=0.14.0 pandas>=2.0.3 plotly>=5.9.0 networkx>=3.1

Next, clone the CRISP repository using the git clone command:

git clone git@github.com:Indranil17/CRISP.git

Now, navigate into the CRISP directory and install the package using pip:

cd CRISP
pip install .

Finally, verify that the installation was successful by launching a Python shell in your terminal and importing the CRISP package:

python
>>> import CRISP
>>> CRISP.atom_indices()

If no errors occurred and the CRISP package is successfully imported, your installation is complete and ready to use.

Testing Installation#

To verify that CRISP is working correctly and to run the test suite, install the testing dependencies:

pip install pytest pytest-cov

Run the test suite to ensure all functionality is working:

# Run all tests
pytest CRISP

# Run tests with coverage report
pytest CRISP --cov=CRISP --cov-report=term-missing

# Generate detailed HTML coverage report
pytest CRISP --cov=CRISP --cov-report=html

The tests require trajectory data in the CRISP/data directory. If tests are skipped due to missing data, this is normal behavior when data files are not available.

Expected Output: The test suite should complete with most tests passing. Some tests may be skipped if specific data files are not present, which is acceptable for basic installation verification.