CHI_SQUARE_ANOVA_1.py

import scipy.stats as stats

# Define null and alternative hypothesis
# For example, if we have three groups: A, B, and C
# H0: The means of all three groups are equal
# H1: At least one of the means is different from the others

# Select significance level
alpha = 0.05

# Sample data
A = [4,3,4]
B = [3,2,5]
C = [5,3,6]

# Calculate sum of all observations
T = sum(A) + sum(B) + sum(C)

# Calculate correction factor
nT = len(A) + len(B) + len(C)
CF = T**2 / nT

# Calculate sum of squares total
SST = sum([x**2 for x in A+B+C]) - CF

# Calculate sum of squares between columns
SSB = sum([sum(g)**2/len(g) for g in [A, B, C]]) - CF

# Calculate sum of squares within columns
SSW = SST - SSB

# Calculate mean of squares between groups
MSB = SSB / 2

# Calculate mean of squares within groups
MSW = SSW / 12

# Calculate F value
Fcal = MSB / MSW

# Calculate critical F value
dfnum = 2  # k - 1
dfden = 12 - 2  # nT - k
Fcrit = stats.f.ppf(q=1-alpha, dfn=dfnum, dfd=dfden)

print(Fcal)
print(Fcrit)

# Compare Fcal and Fcrit
if Fcal > Fcrit:
    print("Reject null hypothesis. At least one of the means is different from the others.")
else:
    print("Fail to reject null hypothesis. The means of all groups are equal.")