"""Reproduce the EJIM South Asia pediatric-adult malnutrition analysis.
Core inputs are published, rounded 2022 age-standardized percentages from
the NCD-RisC India country fact sheet. The ±0.5 simulation is a rounding
sensitivity analysis only; it is not a substitute for model uncertainty intervals.
"""
import numpy as np
import pandas as pd
from scipy import stats

df = pd.read_csv("Additional_file_1_analysis_data.csv")
comparisons = [
    ("Boys vs girls thinness","Boys_thinness_pct","Girls_thinness_pct"),
    ("Boys vs girls obesity","Boys_obesity_pct","Girls_obesity_pct"),
    ("Women vs men obesity","Women_obesity_pct","Men_obesity_pct"),
    ("Women vs girls obesity","Women_obesity_pct","Girls_obesity_pct"),
    ("Men vs boys obesity","Men_obesity_pct","Boys_obesity_pct"),
]
for label,a,b in comparisons:
    res=stats.wilcoxon(df[a],df[b],alternative="two-sided",method="auto")
    print(label, "W=",res.statistic,"p=",res.pvalue)
print("Female pediatric-adult obesity Spearman:",stats.spearmanr(df.Girls_obesity_pct,df.Women_obesity_pct))
print("Male pediatric-adult obesity Spearman:",stats.spearmanr(df.Boys_obesity_pct,df.Men_obesity_pct))