Explore Data

EDA

Author

Affiliation

Eric

Ultragenyx

Published

November 20, 2024

Keywords

EDA, Explore

Code

rm(list=ls(all=TRUE))
library(tinytex)
library(ggplot2)
library(plotly)

library(haven)
library(labelled)
library(bslib)
library(DT)
library(r2rtf)
library(metalite)
library(metalite.table1)
library(haven)
library(tidyr)
library(tools)
library(gtsummary)
library(stringr)
library(RColorBrewer)
library(flextable)
library(stringr)
library(tern)
library(tidyverse)
library(vistime)
library(readxl)
library(hrbrthemes)
library(rlang)
library(lubridate)
library(fuzzyjoin)
library(report)
library(nnet)
library(MASS)
library(dplyr)

library(SmartEDA)
library(summarytools)
library(skimr)
library(ggstatsplot)
library(PerformanceAnalytics)
library(dlookr)
library(ISLR) # for the Wage dataset

Code

ggstatsplot::ggbarstats(data=Wage, x=jobclass,y=education, label="both")

Code

dlookr::describe(iris)%>%flextable()

described_variables	n	mean	sd	se_mean	IQR	skewness	kurtosis	p00	p01	p05	p10	p20	p25	p30	p40	p50	p60	p70	p75	p80	p90	p95	p99	p100
Sepal.Length	150	5.843333	0.8280661	0.06761132	1.3	0.3149110	-0.552064	4.3	4.400	4.600	4.8	5.0	5.1	5.27	5.60	5.80	6.10	6.3	6.4	6.52	6.90	7.255	7.700	7.9
Sepal.Width	150	3.057333	0.4358663	0.03558833	0.5	0.3189657	0.228249	2.0	2.200	2.345	2.5	2.7	2.8	2.80	3.00	3.00	3.10	3.2	3.3	3.40	3.61	3.800	4.151	4.4
Petal.Length	150	3.758000	1.7652982	0.14413600	3.5	-0.2748842	-1.402103	1.0	1.149	1.300	1.4	1.5	1.6	1.70	3.90	4.35	4.64	5.0	5.1	5.32	5.80	6.100	6.700	6.9
Petal.Width	150	1.199333	0.7622377	0.06223645	1.5	-0.1029667	-1.340604	0.1	0.100	0.200	0.2	0.2	0.3	0.40	1.16	1.30	1.50	1.8	1.8	1.90	2.20	2.300	2.500	2.5

Code

iris%>%group_by(Species)%>%univar_numeric()

Adding missing grouping variables: `Species`

$statistics
# A tibble: 12 x 10
   described_variables Species        n    na  mean    sd se_mean   IQR skewness
   <chr>               <fct>      <int> <int> <dbl> <dbl>   <dbl> <dbl>    <dbl>
 1 Petal.Length        setosa        50     0 1.46  0.174  0.0246 0.175   0.106 
 2 Petal.Length        versicolor    50     0 4.26  0.470  0.0665 0.600  -0.607 
 3 Petal.Length        virginica     50     0 5.55  0.552  0.0780 0.775   0.549 
 4 Petal.Width         setosa        50     0 0.246 0.105  0.0149 0.1     1.25  
 5 Petal.Width         versicolor    50     0 1.33  0.198  0.0280 0.3    -0.0312
 6 Petal.Width         virginica     50     0 2.03  0.275  0.0388 0.5    -0.129 
 7 Sepal.Length        setosa        50     0 5.01  0.352  0.0498 0.400   0.120 
 8 Sepal.Length        versicolor    50     0 5.94  0.516  0.0730 0.7     0.105 
 9 Sepal.Length        virginica     50     0 6.59  0.636  0.0899 0.675   0.118 
10 Sepal.Width         setosa        50     0 3.43  0.379  0.0536 0.475   0.0412
11 Sepal.Width         versicolor    50     0 2.77  0.314  0.0444 0.475  -0.363 
12 Sepal.Width         virginica     50     0 2.97  0.322  0.0456 0.375   0.366 
# i 1 more variable: median <dbl>

Code

diamonds%>%diagnose_numeric()%>%flextable()

variables	Q1	mean	median	Q3	max	zero	outlier
carat	4.71	5.731157	5.7	6.54	10.74	8	32
depth	4.71	5.731157	5.7	6.54	10.74	8	32
table	4.71	5.731157	5.7	6.54	10.74	8	32
price	4.71	5.731157	5.7	6.54	10.74	8	32
x	4.71	5.731157	5.7	6.54	10.74	8	32
y	4.71	5.731157	5.7	6.54	10.74	8	32
z	4.71	5.731157	5.7	6.54	10.74	8	32

1 normality test

Code

library(moments)


Attaching package: 'moments'

The following objects are masked from 'package:dlookr':

    kurtosis, skewness

The following objects are masked from 'package:PerformanceAnalytics':

    kurtosis, skewness

Code

agostino.test(airquality$Ozone)


    D'Agostino skewness test

data:  airquality$Ozone
skew = 1.2257, z = 4.6564, p-value = 3.219e-06
alternative hypothesis: data have a skewness

Code

library(DataExplorer)
plot_qq(iris,by="Species")

Code

iris%>%group_by(Species)%>%plot_normality(Petal.Length)

Code

library(ggpubr)


Attaching package: 'ggpubr'

The following objects are masked from 'package:flextable':

    border, font, rotate

Code

ggqqplot(iris, "Sepal.Length", facet.by="Species")

Code

iris%>%group_by(Species)%>%normality()%>%mutate_if(is.numeric, ~round(.,3))%>%flextable()

variable	Species	statistic	p_value	sample
Sepal.Length	setosa	0.978	0.460	50
Sepal.Length	versicolor	0.978	0.465	50
Sepal.Length	virginica	0.971	0.258	50
Sepal.Width	setosa	0.972	0.272	50
Sepal.Width	versicolor	0.974	0.338	50
Sepal.Width	virginica	0.967	0.181	50
Petal.Length	setosa	0.955	0.055	50
Petal.Length	versicolor	0.966	0.158	50
Petal.Length	virginica	0.962	0.110	50
Petal.Width	setosa	0.800	0.000	50
Petal.Width	versicolor	0.948	0.027	50
Petal.Width	virginica	0.960	0.087	50

Code

plot_boxplot(iris, by="Species")

Code

ggstatsplot::ggbetweenstats(data=iris, x=Species,y=Sepal.Length, type="np")

Code

correlate(iris, Petal.Width)%>%flextable()

var1	var2	coef_corr
Petal.Width	Sepal.Length	0.8179411
Petal.Width	Sepal.Width	-0.3661259
Petal.Width	Petal.Length	0.9628654

Code

plot_correlate(iris, method="kendall")

Warning in plot_correlate(iris, method = "kendall"): 'plot_correlate' is deprecated.
Use 'plot.correlate' instead.
See help("Deprecated")

Code

ggcorrmat(data=iris)

Code

# ggcorrmat(data=iris, type="np", output="dataframe")%>%mutate_if(is.numeric, ~round(.,2))%>%flextable()

cor(iris[,1:4], method = "spearman") %>%
  as.data.frame() %>%
  tibble::rownames_to_column("Variable") %>%
  dplyr::mutate(across(where(is.numeric), ~round(., 2))) %>%
  flextable::flextable() %>%
  flextable::theme_vanilla()

Variable	Sepal.Length	Sepal.Width	Petal.Length	Petal.Width
Sepal.Length	1.00	-0.17	0.88	0.83
Sepal.Width	-0.17	1.00	-0.31	-0.29
Petal.Length	0.88	-0.31	1.00	0.94
Petal.Width	0.83	-0.29	0.94	1.00

Code

ggscatterstats(data=iris, x=Sepal.Length, y=Sepal.Width, type="np",marginal.type="boxplot")

Registered S3 method overwritten by 'ggside':
  method from  
  +.gg   GGally

`stat_xsidebin()` using `bins = 30`. Pick better value with `binwidth`.
`stat_ysidebin()` using `bins = 30`. Pick better value with `binwidth`.

Code

chart.Correlation(iris%>%dplyr::select(-Species), method="kendall")

Code

ggplot(iris, aes(Sepal.Length, Sepal.Width))+geom_point()+geom_smooth()+facet_wrap(~Species)

`geom_smooth()` using method = 'loess' and formula = 'y ~ x'

Code

plot_na_intersect(airquality)

Code

plot(imputate_na(airquality, Ozone, Temp, method="knn"))

Code

library(performance)

Warning: package 'performance' was built under R version 4.4.2

Code

plot(check_outliers(airquality$Wind, method="zscore"))

Code

check_outliers(airquality$Wind,method="iqr")

2 outliers detected: cases 9, 48.
- Based on the following method and threshold: iqr (2).
- For variable: airquality$Wind.

-----------------------------------------------------------------------------
Outliers per variable (iqr): 

$`airquality$Wind`
   Row Distance_IQR
9    9     1.527977
48  48     1.614060

Code

dlookr::diagnose_outlier(diamonds)%>%flextable()

variables	outliers_cnt	outliers_ratio	outliers_mean	with_mean	without_mean
carat	32	0.05932518	7.239375	5.731157	5.730262
depth	32	0.05932518	7.239375	5.731157	5.730262
table	32	0.05932518	7.239375	5.731157	5.730262
price	32	0.05932518	7.239375	5.731157	5.730262
x	32	0.05932518	7.239375	5.731157	5.730262
y	32	0.05932518	7.239375	5.731157	5.730262
z	32	0.05932518	7.239375	5.731157	5.730262

Code

airquality%>%dplyr::select(Ozone, Wind)%>%plot_outlier()

Code

plots<-diamonds%>%plot_outlier(diamonds%>%diagnose_outlier()%>%dplyr::filter(outliers_ratio>5)%>%dplyr::select(variables)%>%pull())

Citation

BibTeX citation:

@online{2024,
  author = {, Eric},
  title = {Explore {Data}},
  date = {2024-11-20},
  langid = {en}
}

For attribution, please cite this work as:

Eric. 2024. “Explore Data.” ExploreData. November 20, 2024.