from typing import List
import numpy as np
import pandas as pd
from sklearn.utils.multiclass import type_of_target
[docs]def stack_Xs(X_train=None, X_valid=None, X_test=None):
Xs = []
if X_train is not None:
Xs.append(X_train)
if X_valid is not None:
Xs.append(X_valid)
if X_test is not None:
Xs.append(X_test)
assert len(Xs) > 0
if isinstance(Xs[0], pd.DataFrame):
# 如果拼接后Xs的index已经无序了，先设置一个有序的index
check_and_adjust_Xs_index(Xs)
df = pd.concat(Xs, axis=0)
df.sort_index(inplace=True)
return df
elif isinstance(Xs[0], np.ndarray):
return np.vstack(Xs)
[docs]def check_and_adjust_Xs_index(Xs: List[pd.DataFrame]):
indexes = pd.concat([X.index.to_series() for X in Xs], ignore_index=True).sort_values()
range_index = pd.RangeIndex(indexes.size)
if not np.all(indexes.values == range_index.values):
start = 0
for df in Xs:
df.index = pd.RangeIndex(start, start + df.shape[0])
start += df.shape[0]
return False
return True
[docs]def get_categorical_features_indices(X, columns_metadata):
if isinstance(X, pd.DataFrame):
X = X.values
categorical_features_indices = []
for i in range(X.shape[1]):
col = X[:, i]
try:
col = col.astype("float")
except Exception:
pass
if type_of_target(col) in ("binary", "multiclass"): # todo: debug
categorical_features_indices.append(i)
return categorical_features_indices