PrettyR

Functions to make pretty graphs and tables

To install latest release from Github:

#install.packages("devtools")
devtools::install_github("Jadamso/PrettyR",subdir="PrettyR")
citation("PrettyR")

Example of Tables for non-standed objects

library(PrettyR)

#--------------------------------------------------------------
# Setup

k <- 5
n <- 2000

X <- sapply(seq(k), function(i) runif(n)*i)
colnames(X) <- paste0("X", seq(k))

#FE <- factor(rep( LETTERS[1:4], n/4 ) )
FE <- factor( round(rowSums(X) ) )


Y <- X %*% c(1,2,3,4,5)^1/2 + as.numeric(FE) + rnorm(n)
colnames(Y) <- "Y"

DF <- data.frame(Y,X, FE)

#--------------------------------------------------------------
# Regressions

FORMULAS <- c("Y~X1", "Y~X1+FE",
    "Y~X1+X2+X3+X4+X5", "Y~X1+X2+X3+X4+X5+FE")
REGS <- lapply(FORMULAS, function(form){
    reg <- lm(form, DF)  })

TESTS <- lapply(REGS, function(reg){
    lmtest::coeftest(reg)   })
        
## REGS <- STrollR::mfxall.mc(FORMULAS,1) for vcovSCL se's

#--------------------------------------------------------------
# Table Formatting

new.align <- 'l ll @{\\\\hspace{18pt}}ll'
new.notes <- "All the variables are made up.
    Remember that no statistical procedure is a panacea.
    Table notes should make the table readable by itself.
    And thus may be quite quite quite long."

new.title <- c(" & 1 & 2 & Cat & Cat")
new.tail  <- paste0( paste( c("FE","N","Y","N","Y"), collapse=" & "), " \\\\")
new.metatitle <- c( rep("Type 1",2), rep("Type 2",2) )


stargazerJlist(
    TESTS,
    keep="X1",
    title=new.title,
    label="tab:new",
    column.sep.width="-4pt",
    new.notes=new.notes,
    new.tail=new.tail,
    new.align=new.align)

#--------------------------------------------------------------
# Plotting

scatter_gram(DF$X1, DF$Y)

polygon_plot(REGS[[1]], "X1")

# rci <- polygon_ci(REGS[[2]], "X1")
# polygon_add(rci$X, rci$ci_lu, col=rgb(0,0,1,.25) )
## Looks like noise here because of prediction `type`

Todo

add https://github.com/kklot/R-Utils/blob/master/s2n.R

Caveat Emptor

This is a pre-release to the public and not gauranteed to work correctly with your code and data.