Skip to content

commands.md

Latest commit

 

History

History
476 lines (379 loc) · 19.1 KB

commands.md

File metadata and controls

476 lines (379 loc) · 19.1 KB
Raw

statcpp CLI Command Reference

Basic syntax:

statcpp <category> <command> [options] [file]
statcpp <shortcut> [options] [file]

Input methods:

# 1. Specify a CSV file
statcpp desc mean test/e2e/data/basic.csv --col value

# 2. Pipe from stdin (vertical, one value per line)
cat test/e2e/data/basic.csv | statcpp desc mean --col value

# 3. --row: Pass inline values directly (comma/space-separated)
echo '1,2,3,4,5' | statcpp desc mean --noheader --col 1 --row
echo '1 2 3 4 5' | statcpp desc mean --noheader --col 1 --row
Input Method File Argument Additional Flags Data Format
CSV file Required Not required Standard CSV/TSV
stdin pipe Not required Not required Vertical (one value per line)
--row Not required --noheader --col 1 --row Horizontal (comma/space-separated)

Common Options

Options available for all commands.

Input Control

Flag Default Description
--delimiter <char> Auto-detect CSV delimiter character
--header / --noheader --header Whether a header row is present
--na <strings> NA,NaN,nan,N/A,n/a Strings treated as missing values
--skip_na / --noskip_na --skip_na Exclude missing values from calculations
--fail_na false Exit with error if missing values are present
--presorted false Data is already sorted (skip sorting)
--row false Treat horizontal data as a single column (comma/space-separated)

Output Control

Flag Default Description
--json false Output in JSON format
--quiet false Output numbers only (for piping)

Statistics Common

Flag Default Description
--alpha 0.05 Significance level
--level 0.95 Confidence level
--population false Population statistics (ddof=0)

desc - Descriptive Statistics

Computes descriptive statistics for a single column of numeric data.

Subcommand --col Additional Options Description
summary 1 column --presorted Summary statistics (count, mean, SD, five-number summary, skewness, kurtosis)
mean 1 column Arithmetic mean
median 1 column --presorted Median
mode 1 column Mode (displays all if multiple)
var 1 column --population Variance (default: sample variance)
sd 1 column --population Standard deviation
range 1 column Range (max - min)
iqr 1 column --presorted Interquartile range
cv 1 column Coefficient of variation (SD / Mean)
skewness 1 column Skewness
kurtosis 1 column Excess kurtosis (0 for normal distribution)
percentile 1 column --p (required), --presorted Percentile value
quartiles 1 column --presorted Quartiles (Q1, Q2, Q3)
five-number 1 column --presorted Five-number summary
gmean 1 column Geometric mean
hmean 1 column Harmonic mean
trimmed-mean 1 column --trim (default: 0.1) Trimmed mean (removes upper and lower extremes) 
statcpp desc summary test/e2e/data/basic.csv --col value
statcpp desc mean test/e2e/data/basic.csv --col value
statcpp desc median test/e2e/data/basic.csv --col value
statcpp desc mode test/e2e/data/basic.csv --col value
statcpp desc var test/e2e/data/basic.csv --col score --population
statcpp desc sd test/e2e/data/basic.csv --col score
statcpp desc range test/e2e/data/basic.csv --col value
statcpp desc iqr test/e2e/data/basic.csv --col value
statcpp desc cv test/e2e/data/basic.csv --col value
statcpp desc skewness test/e2e/data/basic.csv --col value
statcpp desc kurtosis test/e2e/data/basic.csv --col value
statcpp desc percentile test/e2e/data/basic.csv --col value --p 0.95
statcpp desc quartiles test/e2e/data/basic.csv --col value
statcpp desc five-number test/e2e/data/basic.csv --col value
statcpp desc gmean test/e2e/data/basic.csv --col value
statcpp desc hmean test/e2e/data/basic.csv --col value
statcpp desc trimmed-mean test/e2e/data/basic.csv --col value --trim 0.1

test - Statistical Tests

You can specify --alternative (two-sided / less / greater) and --alpha (default 0.05).

Subcommand --col Additional Required Options Additional Optional Options Description
t 1 or 2 columns --mu0, --paired, --alternative t-test (auto-detected by number of columns: 1 column = one-sample, 2 columns = two-sample)
welch 2 columns --alternative Welch's t-test
z 1 column --sigma --mu0, --alternative z-test (requires known population standard deviation)
f 2 columns --alternative F-test (test for equality of variances)
shapiro 1 column Shapiro-Wilk normality test
ks 1 column Lilliefors normality test
mann-whitney 2 columns --alternative Mann-Whitney U test
wilcoxon 1 or 2 columns --mu0, --alternative Wilcoxon signed-rank test (1 column = one-sample, 2 columns = paired)
kruskal 2+ columns Kruskal-Wallis test
levene 2+ columns Levene's test (homogeneity of variances)
bartlett 2+ columns Bartlett's test (homogeneity of variances)
chisq 1 or 2 columns Chi-squared test (1 column = goodness-of-fit, 2 columns = observed/expected) 
statcpp test t test/e2e/data/two_groups.csv --col group1 --mu0 25
statcpp test t test/e2e/data/two_groups.csv --col group1,group2
statcpp test t test/e2e/data/two_groups.csv --col group1,group2 --paired
statcpp test welch test/e2e/data/two_groups.csv --col group1,group2
statcpp test z test/e2e/data/two_groups.csv --col group1 --mu0 25 --sigma 3
statcpp test f test/e2e/data/two_groups.csv --col group1,group2
statcpp test shapiro test/e2e/data/two_groups.csv --col group1
statcpp test ks test/e2e/data/two_groups.csv --col group1
statcpp test mann-whitney test/e2e/data/two_groups.csv --col group1,group2
statcpp test wilcoxon test/e2e/data/two_groups.csv --col group1 --mu0 25
statcpp test wilcoxon test/e2e/data/two_groups.csv --col group1,group2
statcpp test kruskal test/e2e/data/scores.csv --col math,science,english
statcpp test levene test/e2e/data/scores.csv --col math,science,english
statcpp test bartlett test/e2e/data/scores.csv --col math,science,english
statcpp test chisq test/e2e/data/basic.csv --col value
statcpp test chisq test/e2e/data/two_groups.csv --col group1,group2

corr - Correlation & Covariance

Subcommand --col Additional Options Description
pearson 2 columns Pearson correlation coefficient
spearman 2 columns Spearman rank correlation coefficient
kendall 2 columns Kendall rank correlation coefficient
cov 2 columns --population Covariance
matrix 2+ columns Correlation matrix 
statcpp corr pearson test/e2e/data/two_groups.csv --col group1,group2
statcpp corr spearman test/e2e/data/two_groups.csv --col group1,group2
statcpp corr kendall test/e2e/data/two_groups.csv --col group1,group2
statcpp corr cov test/e2e/data/two_groups.csv --col group1,group2
statcpp corr matrix test/e2e/data/scores.csv --col math,science,english

effect - Effect Size

Subcommand --col Additional Required Options Additional Optional Options Description
cohens-d 1 or 2 columns --mu0 Cohen's d (1 column = one-sample, 2 columns = two-sample)
hedges-g 1 or 2 columns --mu0 Hedges' g (with small-sample correction)
glass-delta 2 columns Glass's delta (standardized by control group SD)
cohens-h Not required (no CSV input) --p1, --p2 Cohen's h (effect size for proportions)
odds-ratio 1 column (4 values) Odds ratio (4 values: a, b, c, d)
risk-ratio 1 column (4 values) Risk ratio (4 values: a, b, c, d) 
statcpp effect cohens-d test/e2e/data/two_groups.csv --col group1 --mu0 25
statcpp effect cohens-d test/e2e/data/two_groups.csv --col group1,group2
statcpp effect hedges-g test/e2e/data/two_groups.csv --col group1,group2
statcpp effect glass-delta test/e2e/data/two_groups.csv --col group1,group2
statcpp effect cohens-h --p1 0.6 --p2 0.4
statcpp effect odds-ratio test/e2e/data/contingency.csv --col value
statcpp effect risk-ratio test/e2e/data/contingency.csv --col value

ci - Confidence Intervals

Specify the confidence level with --level (default 0.95). ci prop and ci sample-size do not require CSV input.

Subcommand --col Additional Required Options Additional Optional Options Description
mean 1 column --sigma Confidence interval for the mean (z-based if --sigma is specified, t-based otherwise)
diff 2 columns Confidence interval for the difference of means
prop Not required --successes, --trials Confidence interval for a proportion (Wilson)
var 1 column Confidence interval for the variance
sample-size Not required --moe --sigma Required sample size (for means if --sigma is specified, for proportions otherwise) 
statcpp ci mean test/e2e/data/two_groups.csv --col group1
statcpp ci mean test/e2e/data/two_groups.csv --col group1 --level 0.99
statcpp ci mean test/e2e/data/two_groups.csv --col group1 --sigma 3
statcpp ci diff test/e2e/data/two_groups.csv --col group1,group2
statcpp ci prop --successes 45 --trials 100
statcpp ci var test/e2e/data/two_groups.csv --col group1
statcpp ci sample-size --moe 0.03
statcpp ci sample-size --moe 5 --sigma 20

reg - Regression Analysis

The last column in --col is the response variable; the remaining columns are predictor variables.

Subcommand --col Description
simple 2 columns (x, y) Simple linear regression
multiple 3+ columns (x1,...,xp, y) Multiple linear regression
predict 2 columns (x, y) Predicted values
residuals 2 columns (x, y) Residual diagnostics
vif 2+ columns (predictors only) Variance inflation factor (multicollinearity diagnostics) 
statcpp reg simple test/e2e/data/two_groups.csv --col group1,group2
statcpp reg multiple test/e2e/data/scores.csv --col math,science,english
statcpp reg predict test/e2e/data/two_groups.csv --col group1,group2
statcpp reg residuals test/e2e/data/two_groups.csv --col group1,group2
statcpp reg vif test/e2e/data/two_groups.csv --col group1,group2

anova - Analysis of Variance

Each column is treated as one group.

Subcommand --col Description
oneway 2+ columns One-way ANOVA
posthoc-tukey 2+ columns Tukey HSD post hoc test
posthoc-bonferroni 2+ columns Bonferroni post hoc test
posthoc-scheffe 2+ columns Scheffe post hoc test
eta-squared 2+ columns Effect size (eta-squared, omega-squared, Cohen's f) 
statcpp anova oneway test/e2e/data/scores.csv --col math,science,english
statcpp anova posthoc-tukey test/e2e/data/scores.csv --col math,science,english
statcpp anova posthoc-bonferroni test/e2e/data/scores.csv --col math,science,english
statcpp anova posthoc-scheffe test/e2e/data/scores.csv --col math,science,english
statcpp anova eta-squared test/e2e/data/scores.csv --col math,science,english

resample - Resampling

Output depends on random numbers, so results will vary between runs.

Subcommand --col Additional Options Description
bootstrap-mean 1 column Bootstrap confidence interval for the mean
bootstrap-median 1 column Bootstrap confidence interval for the median
bootstrap-sd 1 column Bootstrap confidence interval for the standard deviation
bca 1 column BCa method (bias-corrected and accelerated bootstrap)
permtest 2 columns --paired Permutation test (independent two-sample / paired)
permtest-corr 2 columns Permutation test for correlation 
statcpp resample bootstrap-mean test/e2e/data/basic.csv --col value
statcpp resample bootstrap-median test/e2e/data/basic.csv --col value
statcpp resample bootstrap-sd test/e2e/data/basic.csv --col value
statcpp resample bca test/e2e/data/basic.csv --col value
statcpp resample permtest test/e2e/data/two_groups.csv --col group1,group2
statcpp resample permtest test/e2e/data/two_groups.csv --col group1,group2 --paired
statcpp resample permtest-corr test/e2e/data/two_groups.csv --col group1,group2

ts - Time Series Analysis

Subcommand --col Description
acf 1 column Autocorrelation function (max lag 20)
pacf 1 column Partial autocorrelation function
ma 1 column Moving average (window = 3)
ema 1 column Exponential moving average (alpha = 0.3)
diff 1 column Differencing (first-order)
mae 2 columns (actual, predicted) Mean absolute error
rmse 2 columns (actual, predicted) Root mean squared error
mape 2 columns (actual, predicted) Mean absolute percentage error 
statcpp ts acf test/e2e/data/two_groups.csv --col group1
statcpp ts pacf test/e2e/data/two_groups.csv --col group1
statcpp ts ma test/e2e/data/two_groups.csv --col group1
statcpp ts ema test/e2e/data/two_groups.csv --col group1
statcpp ts diff test/e2e/data/two_groups.csv --col group1
statcpp ts mae test/e2e/data/forecast.csv --col actual,predicted
statcpp ts rmse test/e2e/data/forecast.csv --col actual,predicted
statcpp ts mape test/e2e/data/forecast.csv --col actual,predicted

robust - Robust Statistics

Subcommand --col Additional Options Description
mad 1 column Median absolute deviation (MAD)
outliers 1 column Outlier detection (IQR method)
outliers-zscore 1 column Outlier detection (Z-score method)
outliers-modified 1 column Outlier detection (modified Z-score / MAD-based)
winsorize 1 column --trim (default: 0.1) Winsorization
hodges-lehmann 1 column Hodges-Lehmann estimator
biweight 1 column Biweight midvariance 
statcpp robust mad test/e2e/data/two_groups.csv --col group1
statcpp robust outliers test/e2e/data/basic.csv --col value
statcpp robust outliers-zscore test/e2e/data/basic.csv --col value
statcpp robust outliers-modified test/e2e/data/basic.csv --col value
statcpp robust winsorize test/e2e/data/basic.csv --col value --trim 0.05
statcpp robust hodges-lehmann test/e2e/data/two_groups.csv --col group1
statcpp robust biweight test/e2e/data/two_groups.csv --col group1

survival - Survival Analysis

Event column: 1 = event occurred, 0 = censored.

Subcommand --col Description
kaplan-meier 2 columns (time, event) Kaplan-Meier survival curve
logrank 4 columns (time1, event1, time2, event2) Log-rank test (two-group comparison)
nelson-aalen 2 columns (time, event) Nelson-Aalen cumulative hazard 
statcpp survival kaplan-meier test/e2e/data/survival.csv --col time,event
statcpp survival logrank test/e2e/data/survival_two.csv --col time1,event1,time2,event2
statcpp survival nelson-aalen test/e2e/data/survival.csv --col time,event

cluster - Clustering

Requires 2 or more columns of numeric data. Output may depend on random numbers.

Subcommand --col Description
kmeans 2+ columns k-means clustering (k = 3)
hierarchical 2+ columns Hierarchical clustering
silhouette 2+ columns Silhouette analysis (k = 3) 
statcpp cluster kmeans test/e2e/data/scores.csv --col math,science,english
statcpp cluster hierarchical test/e2e/data/scores.csv --col math,science,english
statcpp cluster silhouette test/e2e/data/scores.csv --col math,science,english

multiple - Multiple Testing Correction

Specify a column of p-values.

Subcommand --col Description
bonferroni 1 column Bonferroni correction
holm 1 column Holm-Bonferroni correction
bh 1 column Benjamini-Hochberg (FDR) correction 
statcpp multiple bonferroni test/e2e/data/pvalues.csv --col pvalue
statcpp multiple holm test/e2e/data/pvalues.csv --col pvalue
statcpp multiple bh test/e2e/data/pvalues.csv --col pvalue

power - Power Analysis

No CSV input required. If --n is specified, power is calculated; if omitted, the required sample size is calculated.

Subcommand --col Required Options Optional Options Description
t-one Not required --effect --n, --power (default: 0.8), --alternative Power for one-sample t-test
t-two Not required --effect --n, --power, --ratio (default: 1.0), --alternative Power for two-sample t-test
prop Not required --p1, --p2 --n, --power, --alternative Power for proportion test 
statcpp power t-one --effect 0.5 --n 30
statcpp power t-one --effect 0.5 --power 0.8
statcpp power t-two --effect 0.5 --power 0.8 --ratio 2.0
statcpp power prop --p1 0.3 --p2 0.5 --n 50

glm - Generalized Linear Models

The last column in --col is the response variable.

Subcommand --col Description
logistic 2+ columns (x1,...,xp, y) Logistic regression (y is binary 0/1)
poisson 2+ columns (x1,...,xp, y) Poisson regression (y is count data) 
statcpp glm logistic test/e2e/data/binary.csv --col x1,x2,y
statcpp glm poisson test/e2e/data/count.csv --col x1,x2,y

model - Model Selection

The last column in --col is the response variable. Requires 3 or more columns (at least 2 predictor variables).

Subcommand --col Description
aic 3+ columns Model comparison by AIC
cv 3+ columns Cross-validation (5-fold)
ridge 3+ columns Ridge regression (lambda = 1.0)
lasso 3+ columns LASSO regression (lambda = 1.0) 
statcpp model aic test/e2e/data/scores.csv --col math,science,english
statcpp model cv test/e2e/data/scores.csv --col math,science,english
statcpp model ridge test/e2e/data/scores.csv --col math,science,english
statcpp model lasso test/e2e/data/scores.csv --col math,science,english

Shortcuts

Frequently used commands can be entered without the category. If the first argument matches a shortcut name, it is automatically expanded to the corresponding category and command.

# The following two are equivalent
statcpp mean data.csv --col value
statcpp desc mean data.csv --col value
Shortcut Expands To
mean desc mean
median desc median
mode desc mode
sd desc sd
var desc var
summary desc summary
range desc range
iqr desc iqr
cv desc cv
skewness desc skewness
kurtosis desc kurtosis
quartiles desc quartiles
gmean desc gmean
hmean desc hmean
ttest test t
pearson corr pearson
spearman corr spearman
kendall corr kendall