Excel weighted stats

Skill cxcscmu/SkillLearnBench/skills/b4-skill-creator-claude-opus-4-6/weighted-gdp-calculation/excel-weighted-stats

How to calculate weighted means and descriptive statistics (min, max, median, percentiles, simple mean) in Excel formulas via openpyxl. Use this skill whenever the user asks for SUMPRODUCT-based weighted averages, GDP-weighted means, percentile calculations, or descriptive statistics in Excel.From its SKILL.md

Install

npx -y skills add cxcscmu/SkillLearnBench --skill excel-weighted-stats

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

2.4 KB, 742 tokens by cl100k_base, as published. Nobody here has run it

Weighted Mean and Descriptive Statistics in Excel

Weighted Mean with SUMPRODUCT

The GDP-weighted mean of a percentage series:

=ROUND(SUMPRODUCT(value_range, weight_range) / SUM(weight_range), 1)

value_range: the calculated percentages (e.g., H35:H40)
weight_range: the GDP values used as weights (e.g., H26:H31)

Both ranges must be the same size. SUMPRODUCT multiplies element-wise and sums.

Why This Works

If pct_i = (exports_i - imports_i) / gdp_i * 100, then:

SUMPRODUCT(pct, gdp) / SUM(gdp)
= SUM(pct_i * gdp_i) / SUM(gdp_i)
= 100 * SUM(net_exports_i) / SUM(gdp_i)

This is the aggregate net-exports-to-GDP ratio for the group.

Descriptive Statistics Formulas

Statistic	Excel Formula
Min	`=MIN(range)`
Max	`=MAX(range)`
Median	`=MEDIAN(range)`
Simple mean	`=AVERAGE(range)`
25th percentile	`=PERCENTILE(range, 0.25)`
75th percentile	`=PERCENTILE(range, 0.75)`

Rounding to 1 Decimal (percentage display)

Wrap in ROUND: =ROUND(MIN(H35:H40), 1)

Net Exports as Percent of GDP

=ROUND((Exports - Imports) / GDP * 100, 1)

Example cell formula:

=ROUND((H12-H19)/H26*100,1)

This yields a value like 12.3 (not 0.123), matching the requirement to display percentages as whole-number-scale values.

openpyxl Example

ws = wb['Task']

# Net exports % of GDP (row 35, columns H-L)
for col in ['H','I','J','K','L']:
    ws[f'{col}35'] = f'=ROUND(({col}12-{col}19)/{col}26*100,1)'

# Descriptive stats (rows 42-47)
for col in ['H','I','J','K','L']:
    ws[f'{col}42'] = f'=ROUND(MIN({col}35:{col}40),1)'
    ws[f'{col}43'] = f'=ROUND(MAX({col}35:{col}40),1)'
    ws[f'{col}44'] = f'=ROUND(MEDIAN({col}35:{col}40),1)'
    ws[f'{col}45'] = f'=ROUND(AVERAGE({col}35:{col}40),1)'
    ws[f'{col}46'] = f'=ROUND(PERCENTILE({col}35:{col}40,0.25),1)'
    ws[f'{col}47'] = f'=ROUND(PERCENTILE({col}35:{col}40,0.75),1)'

# Weighted mean (row 50)
for col in ['H','I','J','K','L']:
    ws[f'{col}50'] = f'=ROUND(SUMPRODUCT({col}35:{col}40,{col}26:{col}31)/SUM({col}26:{col}31),1)'

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.