rf = RFScore() TypeError: 'module' object is not callable

[SanFran-Me]

import os
import pandas as pd
import csv
import oddt
from oddt.scoring.functions import RFScore

oddt.toolkit = oddt.toolkits.rdk

rf = rfscore(version=1, n_jobs=-1)
# rf.load('/path/to/your/model.pkl')

with open('result_csv.csv', 'w', newline='') as csvfile:
    fieldnames = ['pdb_id', 'rf_score_v1']
    writer = csv.DictWriter(csvfile, fieldnames=fieldnames)
    writer.writeheader()

    for root, dirs, _ in os.walk('./coreset'):
        for dir in dirs:
            pdb_id = dir 
            ligands_file = os.path.join(root, dir, f'{pdb_id}_ligand.sdf')
            proteins_file = os.path.join(root, dir, f'{pdb_id}_pocket.pdb')
            lig_mol = next(oddt.toolkit.readfile('sdf', ligands_file))
            prot_mol = next(oddt.toolkit.readfile('pdb', proteins_file))
            score = rf.predict(lig_mol, prot_mol)
            print(f'{pdb_id}_ligand.sdf', score)
            writer.writerow({'pdb_id': pdb_id, 'rf_score_v1': score})
ref_df = pd.read_csv("2016_core_affinity.csv", sep='\t')
result_df = pd.read_csv("result_csv.csv")

BUT IT RAISES ERROR:

Traceback (most recent call last):
  File "oddt_rfscore_Coreset.py", line 9, in <module>
    rf = RFScore(version=1, n_jobs=-1)
TypeError: 'module' object is not callable

[SanFran-Me]

I saw the source code od RFScore:

def load(self, filename=None, version=1, pdbbind_version=2016):
        if filename is None:
            fname = 'RFScore_v%i_pdbbind%i.pickle' % (version, pdbbind_version)

BUT I did not find any RFScore_v?_pdbbind?.pickle...
So if I want to use the default pkl, what do I need to do?

[RJ-Li]

I also do not know how to train models on CASF-2016 refined-set and get its pkl model file(for further re-training).
The only thing I know is that there is a csv containing affinity and descriptors in oddt-master/oddt/scoring/functions/RFScore/rfscore_descs_v1.csv

And when I tried:

rf = RFScore.load('/path/to/your/model.pkl')
AttributeError: module 'oddt.scoring.functions.RFScore' has no attribute 'load'