Stella

A package of functions that mimic some basic Stata functionality.

Installation

  ] add https://github.com/mwsohn/Stella.jl

List of Functions

Here's the list of functions. Documentation for individual functions can be accessed using the Julia help (e.g., ?read_stata at the Julia REPL prompt).

DataFrame tools

• acompress - Compress a vector
• classify - Creates a classification variable with the specified cutoff values
• descr - Prints the list of variables, similar to showcols
• dfcompress - Compress all numeric variables in a dataframe by converting them to the type that will occupy the least memory.
• dfmerge - Merges two dataframes and creates a :_merge variable that indicates data sources, similar to Stata merge command
• dfsample - Creates a dataframe with a randomly selected sample from the input dataframe
• duplicates - Identifies, deletes, or tag duplicates in a dataframe
• keepfirst - Keep the first n rows from the same group in grouped data (default: n = 1)
• keeplast - Keep the last n rows from the same group in grouped data (default: n = 1)
• pickone - Identifies only one record in repeated data similar to egen tag function in Stata
• renvars! - Change case for all variables (default: )
• xtile - Creates a ntile classification variable, similar to Stata's xtile command

Row functions that work on a column of values in a subgroup and attaches return values to the original DataFrame

• rowfirst - Returns the first non-missing value
• rowlast - Returns the last non-missing value
• rowmax - Returns the maximum non-missing value
• rowmin - Returns the minimum non-missing value
• rowmean - Returns the mean of non-missing values
• rowmedian - Returns the median of non-missing values
• rowmiss - Returns the number of missing values
• rownonmiss - Returns the number of non-missing values
• rowsd - Returns the standard deviation of non-missing values
• rowpctile - Returns the percentile value of non-missing values (default: p = 0.5)
• rowtotal - Returns the sum of non-missing values

Sub functions that work on a subgroup vector in a grouped data. A variable is automatically created with "_total", "_mean", "_sd", etc, appended after the statvar.

• subfirst! - Returns the first non-missing value
• sublast! - Returns the last non-missing value
• submax! - Returns the maximum non-missing value
• submin! - Returns the minimum non-missing value
• submean! - Returns the mean of non-missing values
• submedian! - Returns the median of non-missing values
• submiss! - Returns the number of missing values
• subnonmiss! - Returns the number of non-missing values
• subsd! - Returns the standard deviation of non-missing values
• subpctile! - Returns the percentile value of non-missing values (default: p = 0.5)
• subtotal! - Returns the sum of non-missing values

Statitical utilities for interactive use

• pwcorr - Produces pairwise correlations
• univ - Univariate statistics, output to a DataFrame
• ttest - T-test statistics
• tabstat - Creates statistics by categories using functions provided, similar to Stata's tabstat command
• eform - Produces exponentiated GLM regression output such as coefficients and standard errors