File size: 6,609 Bytes
798602c | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 | from __future__ import annotations
from typing import Iterable, Tuple
import numpy as np
import pandas as pd
from core.hypothesis_tests import (
one_sample_ttest,
two_sample_ttest,
variance_test,
one_way_anova,
)
ROUND = 4
def _round_table(table: pd.DataFrame, decimals: int = ROUND) -> pd.DataFrame:
"""Round only numeric columns of the result table."""
if table is None:
return table
tbl = table.copy()
num_cols = tbl.select_dtypes(include="number").columns
if len(num_cols) > 0:
tbl[num_cols] = tbl[num_cols].round(decimals)
return tbl
def _ensure_numeric_series(df: pd.DataFrame, column: str) -> np.ndarray:
if df is None:
raise ValueError("No dataset loaded.")
if column not in df.columns:
raise ValueError(f"Column '{column}' not found in the dataset.")
series = df[column].dropna()
if series.empty:
raise ValueError("No valid data in the selected column.")
return series.to_numpy()
def _materialize_group(
df: pd.DataFrame,
numeric_col: str,
cat_col: str | None,
cat_vals: Iterable[str],
) -> np.ndarray:
if cat_col is None:
raise ValueError("No categorical column selected.")
if cat_col not in df.columns:
raise ValueError(f"Categorical column '{cat_col}' not found in the dataset.")
# Cast selected values to the actual dtype of the column
if cat_vals is None:
values = []
else:
values = list(cat_vals)
if not values:
raise ValueError(f"No categories selected for column '{cat_col}'.")
cat_series = pd.Series(values).astype(df[cat_col].dtype)
mask = df[cat_col].isin(cat_series)
series = df.loc[mask, numeric_col].dropna()
if series.empty:
raise ValueError("One or more groups are empty after filtering.")
return series.to_numpy()
def run_hypothesis_testing(
*,
df: pd.DataFrame | None,
numeric_col: str,
hypo_test: str,
mu0_text: str,
alternative: str,
include_graph: bool,
bootstrap_samples: int,
cat_col1: str | None,
cat_vals1: list[str],
name_group1: str,
cat_col2: str | None,
cat_vals2: list[str],
name_group2: str,
cat_col3: str | None,
cat_vals3: list[str],
plot_type: str,
correction: bool,
test_type: str,
) -> Tuple[pd.DataFrame, object | None]:
"""
High-level dispatcher used by the Hypothesis Testing tab.
Returns:
(result_table, figure_or_none)
"""
if df is None:
raise ValueError("No dataset loaded.")
# Common numeric data check
_ = _ensure_numeric_series(df, numeric_col)
# ------------------------------------------------------------
# One-sample t-test
# ------------------------------------------------------------
if hypo_test == "One sample Student's t-test":
if not mu0_text.strip():
raise ValueError("μ₀ must be specified for the one-sample t-test.")
try:
mu0 = float(mu0_text)
except Exception:
raise ValueError("μ₀ must be a numeric value.")
sample = df[numeric_col].dropna().to_numpy()
table, fig = one_sample_ttest(
sample=sample,
mu0=mu0,
alternative=alternative,
numeric_col=numeric_col,
bootstrap_samples=bootstrap_samples,
include_graph=include_graph,
)
table = _round_table(table)
return table, fig
# ------------------------------------------------------------
# Two-sample t-test
# ------------------------------------------------------------
if hypo_test == "Two samples Student's t-test":
group1 = _materialize_group(df, numeric_col, cat_col1, cat_vals1)
group2 = _materialize_group(df, numeric_col, cat_col2, cat_vals2)
# If names are empty, fall back to defaults
name1 = name_group1 or "Group 1"
name2 = name_group2 or "Group 2"
table, fig = two_sample_ttest(
group1=group1,
group2=group2,
numeric_col=numeric_col,
name_group1=name1,
name_group2=name2,
alternative=alternative,
correction=correction,
plot_type=plot_type,
bootstrap_samples=bootstrap_samples,
include_graph=include_graph,
)
table = _round_table(table)
return table, fig
# ------------------------------------------------------------
# Equal variance between two groups
# ------------------------------------------------------------
if hypo_test == "Equal variance between two groups":
group1 = _materialize_group(df, numeric_col, cat_col1, cat_vals1)
group2 = _materialize_group(df, numeric_col, cat_col2, cat_vals2)
name1 = name_group1 or "Group 1"
name2 = name_group2 or "Group 2"
table, fig = variance_test(
group1=group1,
group2=group2,
name_group1=name1,
name_group2=name2,
test_type=test_type,
include_graph=include_graph,
bootstrap_samples=bootstrap_samples,
)
table = _round_table(table)
return table, fig
# ------------------------------------------------------------
# One-way ANOVA
# ------------------------------------------------------------
if hypo_test == "One-way ANOVA":
if cat_col3 is None:
raise ValueError("A categorical column must be selected for ANOVA.")
if cat_col3 not in df.columns:
raise ValueError(
f"Categorical column '{cat_col3}' not found in the dataset."
)
if not cat_vals3:
raise ValueError("At least one category must be selected for ANOVA.")
cat_series = pd.Series(cat_vals3).astype(df[cat_col3].dtype)
data_group = df[df[cat_col3].isin(cat_series)][[numeric_col, cat_col3]].dropna()
table, fig = one_way_anova(
data_group=data_group,
numeric_col=numeric_col,
cat_col=cat_col3,
)
table = _round_table(table)
return table, fig
# ------------------------------------------------------------
# Fallback
# ------------------------------------------------------------
raise ValueError(f"Unknown hypothesis test: {hypo_test}")
|