tomasokal / phackr

Package to produce tables of directionality and significance of ologit regressions ran across multiple dependent variables. Useful for checking significance of a battery of questions

phackr

The goal of phackr is to produce excel sheets showing multivariate significance for several dependent variables across as many covariates as specified. The function currently only supports ordered logistic regression models. The future plan for the package is to allow specification of several interchangeable covariates to output to multiple excel sheets.

Installation

You can install the development version from github with:

library(devtools)
install_github("willdebras/phackr")

library(phackr)
library(tibble)

Example

You can use the phackr_setup() function to setup a survey object

library(rio)
data <- import("full_data.dta")

survey_data <- phackr_setup(data = data, weight = "finalwt", caseid = "su_id")

Once it is setup, you can use the phackr() function to actually produce the tables.

sheet1 <- phackr(data = survey_data, 
                 dvs = c("factor(trade1a)", "factor(trade1b)", "factor(trade1c)", "factor(trade1d)", "factor(trade1e)", "factor(trade1f)", "factor(trade1g)", "factor(trade2)", "factor(trade3a)", "factor(trade3b)", "factor(trade3c)"), 
                 demos = c("urban", "marital", "agegrp", "education", "hhincome", "empstatus", "gender"))

library(knitr)
kable(sheet1)

	factor(trade1a)	factor(trade1b)	factor(trade1c)	factor(trade1d)	factor(trade1e)	factor(trade1f)	factor(trade1g)	factor(trade2)	factor(trade3a)	factor(trade3b)	factor(trade3c)
urban								-		+
marital				-	-						+
agegrp
education			-					+	+	-
hhincome		-
empstatus
gender					-	+		+