causaldag.classes.dag.DAG.from_dataframe

classmethod DAG.from_dataframe(df)[source]

Create a DAG from a dataframe, where the indices and columns are node names and a nonzero entry indicates the presence of an edge.

Parameters:df – The pandas dataframe.
Returns:The graph as a DAG object.
Return type:DAG

Examples

>>> import causaldag as cd
>>> import numpy as np
>>> import pandas as pd
>>> amat = np.array([[0, 1], [0, 0]])
>>> df = pd.DataFrame(amat, index=["a", "b"], columns=["a", "b"])
>>> d = cd.DAG.from_dataframe(df)
>>> d.arcs
{('a', 'b')}