I don't find CWMM formula

eleven_sigma · March 9, 2024, 12:00am

In CWMM are predictions and meta model ranked and gaussianized?
Both use pot = 1.5 too or only preds or none?
I don’t find the code for CWMM in

github.com

numerai/numerai-tools/blob/master/numerai_tools/scoring.py

from typing import List, Tuple, Union, Optional

import numpy as np
import pandas as pd
from scipy import stats
from sklearn.preprocessing import OneHotEncoder


# sometimes when we match up the target/prediction indices,
# changes in Numerai data generation can cause some stocks to get filtered out
# (e.g. when there are a lot of NaNs in the target),
# this ensures we don't filter too much
DEFAULT_MAX_FILTERED_INDEX_RATIO = 0.2


# this is primarily used b/c round 326 had too many stocks,
# so we need to filter out the unnecessary ids here just in case.
# it's also just convenient way to ensure everything is sorted/matching
def filter_sort_index(
    s1: Union[pd.DataFrame, pd.Series],

This file has been truncated. show original

EDIT:

Where is CWMM? is numerai_corr function used between mm and preds?