centrality_encoder.py

import pandas as pd
import networkx as nx
from sklearn.base import BaseEstimator, TransformerMixin


class CentralityEncoder(BaseEstimator, TransformerMixin):
    def __init__(self, node_column, edge_column):
        self.mapping_dict = {}
        self.node_column = node_column
        self.edge_column = edge_column

    def fit(self, X, Y=None):
        G = nx.from_pandas_edgelist(X, self.node_column, self.edge_column)
        centrality = nx.eigenvector_centrality(G)
        self.mapping_dict.update(centrality)
        return self

    def transform(self, X, Y=None):
        X[self.node_column] = X[self.node_column].map(self.mapping_dict)
        X[self.edge_column] = X[self.edge_column].map(self.mapping_dict)
        return X


df = pd.DataFrame({
    'pass': [1, 0, 0, 1, 1, 1, 1, 0, 1, 1],
    'teacher': ['a', 'a', 'b', 'a', 'b', 'b', 'c', 'c', 'a', 'd'],
    'advisor': ['b', 'c', 'c', 'a', 'd', 'd', 'a', 'b', 'a', 'c']
})
print(df)
df = CentralityEncoder(node_column='teacher', edge_column='advisor').fit_transform(df)
print(df)