README.md

In tengku-hanis/niceFunction: These are random nice functions that I found

niceFunction

The goal of niceFunction is to retain all random function that I found all over the books, forums, etc.

Installation

You can install the the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("tengku-hanis/niceFunction")

Example

This is a summary example which shows the use of each function. As of now, this package only have 6 functions:

1. histWithCurve
2. histCurve
3. histNA_byVar
4. regDiag
5. read_excel_allsheets
6. changeType

library(niceFunction)
library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union

histWithCurve give a histogram with a normal density curve.

histWithCurve(iris$Sepal.Length)

histCurve give a ggplot2 histogram with a normal density curve.

histCurve(iris, Sepal.Length)

histNA_byVar assess the distribution of NAs of certain variable is affected by another variable.

dat <- iris
dat[dat$Species == "setosa", "Sepal.Length"] <- NA
histNA_byVar(dat, Sepal.Length, Sepal.Width)

The histogram with NA values (label by True) indicate a right-tailed missingness compared to the histogram with no NAs (label by False).

regDiag is used for screening of outliers and influential cases.

# Create some outlier observations
iris[151, ] <- c(9, 9, 9, 9, "virginica")
iris <- iris %>% 
  mutate(across(c(Sepal.Length:Petal.Width), as.numeric))

mod <- lm(Sepal.Length ~ Species + Sepal.Width, data = iris)
regDiag(mod)
#> 
#> 
#> leverage    Freq
#> ---------  -----
#> FALSE        150
#> TRUE           1
#> 
#> SDR      Freq
#> ------  -----
#> FALSE     148
#> TRUE        3
#> 
#> DFFits    Freq
#> -------  -----
#> FALSE      151
#> 
#> DFBetas    Freq
#> --------  -----
#> FALSE       603
#> TRUE          1
#> 
#> cook.d    Freq
#> -------  -----
#> FALSE      150
#> TRUE         1

True indicate the presence of outliers and/or influential cases according to that metrics and vice-versa.

read_excel_allsheets read all excel sheets or several excel sheets.

## Read all excel sheets (not run)
# read_excel_allsheets("datasets")

## Read several excel sheets (not run)
# read_excel_allsheets("datasets", pages = 2:5)

changeType change the variable type across list of data frames.

# Make a list
iris_list <- list(iris1 = iris, iris2 = iris)

# Change one variable type
iris_list <- lapply(iris_list, changeType, Var = "Sepal.Width", funct = "as.character")

# Change 2 variables type
iris_list <- lapply(iris_list, changeType, Var = c("Sepal.Length", "Species"), funct = "as.character")

tengku-hanis/niceFunction documentation built on Oct. 27, 2021, 7:41 a.m.