Open HJW3536 opened 3 years ago
My advice is to encode the text with a label encoder like this:
le = LabelEncoder()
le.fit(df.stack().unique())
# save label encoder
np.save('label_encoder.npy', le)
df['source'] = le.transform(df['source'])
df['destination'] = le.transform(df['destination'])
then
le = np.load('label_encoder.npy', allow_pickle=True).tolist()
# get node_id
le.transform([node_name])[0]
# get node name
le.inverse_transform([node_id])[0]
Chinese characters represent nodes, and numbers represent weights.Has anyone done a similar experiment? Ask for advice