Source code for autoflow.pipeline.components.preprocessing.operate.split.cat_num

import numpy as np
import pandas as pd

from autoflow.pipeline.components.preprocessing.operate.split.base import BaseSplit
from autoflow.utils.data import is_cat

__all__ = ["SplitCatNum"]


[docs]class SplitCatNum(BaseSplit):
[docs] def calc_R(self, col, rows): return np.unique(col).size / rows
key1_hp_name = "cat_name" key2_hp_name = "num_name" key1_default_name = "cat" key2_default_name = "num" key1 = "cat" key2 = "num"
[docs] def judge_keyname(self, col: pd.Series, rows): if is_cat(col): keyname = self.key1 else: keyname = self.key2 return keyname