SmartEDA

---

Background

In a quality statistical data analysis the initial step has to be exploratory. Exploratory data analysis begins with the univariate exploratory analyis - examining the variable one at a time. Next comes bivariate analysis followed by multivariate analyis. SmartEDA package helps in getting the complete exploratory data analysis just by running the function instead of writing lengthy r code.

Functionalities of SmartEDA

The SmartEDA R package has four unique functionalities as

• Descriptive statistics
• Data visualization
• Custom table
• HTML EDA report

Comparison with other packages

SmartEDA package with other similar packages available in CRAN for exploratory data analysis viz. dlookr, DataExplorer, Hmisc, exploreR, RtutoR and summarytools. The metric for evaluation is the availability of various desired features for performing an Exploratory data analysis

Journal of Open Source Software Article

An article describing SmartEDA package for exploratory data analysis approach has been published in arxiv and Journal of Open Source Software JOSS. Please cite the paper if you use SmartEDA in your work!

Installation

The package can be installed directly from CRAN.

R install.packages("SmartEDA")

You can install the latest development verion of the SmartEDA from github with:

R install.packages("devtools") devtools::install_github("daya6489/SmartEDA",ref = "develop")

Example

Data

In this vignette, we will be using a simulated data set containing sales of child car seats at 400 different stores.

Data Source ISLR package.

Install the package "ISLR" to get the example data set.

R install.packages("ISLR") library("ISLR") install.packages("SmartEDA") library("SmartEDA") ## Load sample dataset from ISLR pacakge Carseats= ISLR::Carseats

Overview of the data

Understanding the dimensions of the data set, variable names, overall missing summary and data types of each variables

```R

overview of the data;

ExpData(data=Carseats,type=1)

structure of the data

ExpData(data=Carseats,type=2)

```

Summary of numerical variables

To summarise the numeric variables, you can use following r codes from this pacakge

```R

Summary statistics by – overall

ExpNumStat(Carseats,by="A",gp=NULL,Qnt=seq(0,1,0.1),MesofShape=2,Outlier=TRUE,round=2)

Summary statistics by – overall with correlation

ExpNumStat(Carseats,by="A",gp="Price",Qnt=seq(0,1,0.1),MesofShape=1,Outlier=TRUE,round=2)

Summary statistics by – category

ExpNumStat(Carseats,by="GA",gp="Urban",Qnt=seq(0,1,0.1),MesofShape=2,Outlier=TRUE,round=2)

```

weighted summary for numerical variables

R ExpNumStat(mtcars,by="A",round=2, weight = "wt")

Graphical representation of all numeric features

```R

Generate Boxplot by category

ExpNumViz(mtcars,target="gear",type=2,nlim=25,fname = file.path(tempdir(),"Mtcars2"),Page = c(2,2))

Generate Density plot

ExpNumViz(mtcars,target=NULL,type=3,nlim=25,fname = file.path(tempdir(),"Mtcars3"),Page = c(2,2))

Generate Scatter plot

ExpNumViz(mtcars,target="carb",type=3,nlim=25,fname = file.path(tempdir(),"Mtcars4"),Page = c(2,2))

```

Summary of Categorical variables

```R

Frequency or custom tables for categorical variables

ExpCTable(Carseats,Target=NULL,margin=1,clim=10,nlim=5,round=2,bin=NULL,per=T)
ExpCTable(Carseats,Target="Price",margin=1,clim=10,nlim=NULL,round=2,bin=4,per=F)
ExpCTable(Carseats,Target="Urban",margin=1,clim=10,nlim=NULL,round=2,bin=NULL,per=F)    

Summary statistics of categorical variables

ExpCatStat(Carseats,Target="Urban",result = "Stat",clim=10,nlim=5,Pclass="Yes")

Inforamtion value and Odds value

ExpCatStat(Carseats,Target="Urban",result = "IV",clim=10,nlim=5,Pclass="Yes")

```

weighted count for categorical variables

R ExpCTable(mtcars, margin = 1, clim = 10, nlim = 3, bin = NULL, per = FALSE, weight = "wt"")

Graphical representation of all categorical variables

```R

column chart

ExpCatViz(Carseats,target="Urban",fname=NULL,clim=10,col=NULL,margin=2,Page = c(2,1),sample=2)

Stacked bar graph

ExpCatViz(Carseats,target="Urban",fname=NULL,clim=10,col=NULL,margin=2,Page = c(2,1),sample=2)

Variable importance graph using information values

ExpCatStat(Carseats,Target="Urban",result="Stat",Pclass="Yes",plot=TURE,top=20,Round=2) ```

Variable importance based on Information value

R ExpCatStat(Carseats,Target="Urban",result = "Stat",clim=10,nlim=5,bins=10,Pclass="Yes",plot=TRUE,top=10,Round=2)

Create HTML EDA report

Create a exploratory data analysis report in HTML format

R ExpReport(Carseats,Target="Urban",label=NULL,op_file="test.html",op_dir=getwd(),sc=2,sn=2,Rc="Yes")

Quantile-quantile plot for numeric variables

R ExpOutQQ(CData,nlim=10,fname=NULL,Page=c(2,2),sample=4) <!-- -->

Parallel Co-ordinate plots

```R

Defualt ExpParcoord funciton

ExpParcoord(CData,Group=NULL,Stsize=NULL,Nvar=c("Price","Income","Advertising","Population","Age","Education"))

With Stratified rows and selected columns only

ExpParcoord(CData,Group="ShelveLoc",Stsize=c(10,15,20),Nvar=c("Price","Income"),Cvar=c("Urban","US"))

Without stratification

ExpParcoord(CData,Group="ShelveLoc",Nvar=c("Price","Income"),Cvar=c("Urban","US"),scale=NULL)

Scale change

ExpParcoord(CData,Group="US",Nvar=c("Price","Income"),Cvar=c("ShelveLoc"),scale="std")

Selected numeric variables

ExpParcoord(CData,Group="ShelveLoc",Stsize=c(10,15,20),Nvar=c("Price","Income","Advertising","Population","Age","Education"))

Selected categorical variables

ExpParcoord(CData,Group="US",Stsize=c(15,50),Cvar=c("ShelveLoc","Urban")) ``` <!-- -->

Two independent plots side by side for the same variable

To plot graph from same variable when Target=NULL vs. when Target = categorical variable (binary or multi-class variable)

```R target = "gear" categoricalfeatures <- c("vs", "am", "carb") numeircalfeatures <- c("mpg", "cyl", "disp", "hp", "drat", "wt", "qsec")

num1 <- ExpTwoPlots(mtcars, plottype = "numeric", ivvariables = numeircalfeatures, target = "gear", lparglist = list(alpha=0.5, color = "red", fill= "white", binwidth=1), lpgeomtype = 'histogram', rparglist = list(alpha=0.5, fill = c("red", "orange", "pink"), binwidth=1), rpgeomtype = 'histogram', fname = "dub2.pdf", page = c(2,1), theme = "Default")

``` <!-- -->

Univariate Outlier analysis

In statistics, an outlier is a data point that differs significantly from other observations. An outlier may be due to variability in the measurement or it may indicate experimental error; the latter are sometimes excluded from the data set.An outlier can cause serious problems in statistical analyses.

Identifying outliers: There are several methods we can use to identify outliers. In ExpOutliers used two methods (1) Boxplot and (2) Standard Deviation

```R

Identifying outliers mehtod - Boxplot

ExpOutliers(Carseats, varlist = c("Sales","CompPrice","Income"), method = "boxplot", capping = c(0.1, 0.9))

Identifying outliers mehtod - 3 Standard Deviation

ExpOutliers(Carseats, varlist = c("Sales","CompPrice","Income"), method = "3xStDev", capping = c(0.1, 0.9))

Identifying outliers mehtod - 2 Standard Deviation

ExpOutliers(Carseats, varlist = c("Sales","CompPrice","Income"), method = "2xStDev", capping = c(0.1, 0.9))

Create outlier flag (1,0) if there are any outliers

ExpOutliers(Carseats, varlist = c("Sales","CompPrice","Income"), method = "3xStDev", capping = c(0.1, 0.9), outflag = TRUE)

Impute outlier value by mean or median valie

ExpOutliers(Carseats, varlist = c("Sales","CompPrice","Income"), method = "3xStDev", treatment = "mean", capping = c(0.1, 0.9), outflag = TRUE)

```

Exploratory analysis - Custom tables, summary statistics

Descriptive summary on all input variables for each level/combination of group variable. Also while running the analysis we can filter row/cases of the data.

R ExpCustomStat(Carseats,Cvar=c("US","Urban","ShelveLoc"),gpby=FALSE) ExpCustomStat(Carseats,Cvar=c("US","Urban"),gpby=TRUE,filt=NULL) ExpCustomStat(Carseats,Cvar=c("US","Urban","ShelveLoc"),gpby=TRUE,filt=NULL) ExpCustomStat(Carseats,Cvar=c("US","Urban"),gpby=TRUE,filt="Population>150") ExpCustomStat(Carseats,Cvar=c("US","ShelveLoc"),gpby=TRUE,filt="Urban=='Yes' & Population>150") ExpCustomStat(Carseats,Nvar=c("Population","Sales","CompPrice","Income"),stat = c('Count','mean','sum','var','min','max')) ExpCustomStat(Carseats,Nvar=c("Population","Sales","CompPrice","Income"),stat = c('min','p0.25','median','p0.75','max')) ExpCustomStat(Carseats,Nvar=c("Population","Sales","CompPrice","Income"),stat = c('Count','mean','sum','var'),filt="Urban=='Yes'") ExpCustomStat(Carseats,Nvar=c("Population","Sales","CompPrice","Income"),stat = c('Count','mean','sum'),filt="Urban=='Yes' & Population>150") ExpCustomStat(data_sam,Nvar=c("Population","Sales","CompPrice","Income"),stat = c('Count','mean','sum','min'),filt="All %ni% c(999,-9)") ExpCustomStat(Carseats,Nvar=c("Population","Sales","CompPrice","Education","Income"),stat = c('Count','mean','sum','var','sd','IQR','median'),filt=c("ShelveLoc=='Good'^Urban=='Yes'^Price>=150^ ^US=='Yes'")) ExpCustomStat(Carseats,Cvar = c("Urban","ShelveLoc"), Nvar=c("Population","Sales"), stat = c('Count','Prop','mean','min','P0.25','median','p0.75','max'),gpby=FALSE) ExpCustomStat(Carseats,Cvar = c("Urban","US","ShelveLoc"), Nvar=c("CompPrice","Income"), stat = c('Count','Prop','mean','sum','PS','min','max','IQR','sd'), gpby = TRUE) ExpCustomStat(Carseats,Cvar = c("Urban","US","ShelveLoc"), Nvar=c("CompPrice","Income"), stat = c('Count','Prop','mean','sum','PS','P0.25','median','p0.75'), gpby = TRUE,filt="Urban=='Yes'") ExpCustomStat(data_sam,Cvar = c("Urban","US","ShelveLoc"), Nvar=c("Sales","CompPrice","Income"), stat = c('Count','Prop','mean','sum','PS'), gpby = TRUE,filt="All %ni% c(888,999)") ExpCustomStat(Carseats,Cvar = c("Urban","US"), Nvar=c("Population","Sales","CompPrice"), stat = c('Count','Prop','mean','sum','var','min','max'), filt=c("ShelveLoc=='Good'^Urban=='Yes'^Price>=150"))

Articles

See article wiki page.