定義總平方和分解公式:
利用檢驗統計量F定義檢驗方法:
'''實現單因素方差分析'''
# 導入相關包
import pandas as pd
import numpy as np
import math
import scipy
from scipy import stats
# 自定義函數
def SST(Y):
sst = sum(np.power(Y - np.mean(Y), 2))
return sst
def SSA(data, x_name, y_name):
total_avg = np.mean(data[y_name])
df = data.groupby([x_name]).agg(['mean', 'count'])
df = df[y_name]
ssa = sum(df["count"]*(np.power(df["mean"] - total_avg, 2)))
return ssa
def SSE(data, x_name, y_name):
df = data.groupby([x_name]).agg(['mean'])
df = df[y_name]
#dict_ = dict(df["mean"]) 用dict函數報錯
dict_=df["mean"].to_dict()
data_ = data[[x_name, y_name]]
data_["add_mean"] = data_[x_name].map(lambda x: dict_[x])
sse = sum(np.power(data_[y_name] - data_["add_mean"], 2))
return sse
def one_way_anova(data, x_name, y_name, alpha=0.05):
n = len(data) # 總觀測值數
k = len(data[x_name].unique()) # 變量水平個數
sst = SST(data[y_name]) # 總平方和
ssa = SSA(data, x_name, y_name) # 組間平方和
sse = SSE(data, x_name, y_name) # 組內平方和
msa = ssa / (k-1) # 組間均方 或 組間方差
mse = sse / (n-k) # 組內均方 或 組內方差
F = msa / mse # 檢驗統計量F
pf = scipy.stats.f.sf(F, k-1, n-k)
Fa = scipy.stats.f.isf(alpha, dfn=k-1, dfd=n-k) # F臨界值
r_square = ssa / sst # 自變量與因變量的關係強度表示
table = pd.DataFrame({'差異源':['組間', '組內', '總和'],
'平方和SS':[ssa, sse, sst],
'自由度df':[k-1, n-k, n-1],
'均方MS':[msa, mse, '_'],
'F值':[F, '_', '_'],
'P值':[pf, '_', '_'],
'F臨界值':[Fa, '_', '_'],
'R^2':[r_square, '_', '_']})
return table
實例測試結果:
d1=pd.read_excel(r'C:/Users/LHL/Desktop/方差分析.xlsx')
one_way_anova(d1, 'X', 'Y', alpha=0.05)