SMI205 Producing HTML tables v2

Technical notes

• These are guidelines and templates on how to prepare publishable tables in R Markdown printed into HTML format for SMI205 students.

• Before running this R Markdown document install the following packages:

install.packages("rmarkdown")
install.packages("rmdformats")
install.packages("essurvey")
install.packages("xtable")
install.packages("summarytools")
install.packages("knitr")
install.packages("sjPlot")

• This HTML page was produced using the default R Markdown theme and ‘tango’ highlight, with a button to download Rmd code. The same document, but in material theme from rmdformats package is here.

html_document: highlight: tango code_download: true toc: true toc_depth: 2 toc_float: collapsed: false smooth_scroll: true

• More YAML settings are overviewed here: https://rmarkdown.rstudio.com/docs/reference/html_document.html

• I’m also setting the global chunk options, so for all r chunks the code is diplayed (echo=TRUE - note that default in this theme is FALSE), but any additional messages or warnings are not displayed (both set up as FALSE).

library(knitr)
library(rmdformats)
## Global options
options(max.print="85")
opts_chunk$set(echo=TRUE,
                 cache=TRUE,
               prompt=FALSE,
               tidy=TRUE,
               comment=NA,
               message=FALSE,
               warning=FALSE)
opts_knit$set(width=85)

• With these global settings I do not have to specify these options seperately for all my r chunks, only if I want to change them. You might want to specify include=FALSE for this r chunk, so it is not displayed at the begining of your report while the code is still run.

Opening and subsetting data

I use ESS wave 2016 (wave 8) for this excercise. As you remember from our previous R labs, this data can be easily dowloaded using essurvey package.

In the settings of this r chunk I specify results = 'hide' so R Markdown does not print all information about loaded data. Try changing this to see what happens.

I subset data and keep the following variables in my new, smaller dataset called round_8_final: idno, cntry, blgetmg, imbgeco, imueclt, imwbcnt, gndr, agea, maritalb, eisced, polintr, hincfel, uemp5yr.

library("essurvey")
set_email("a.piekut@sheffield.ac.uk")
round_8 <- import_rounds(8)
round_8_final = subset(round_8, select = c(idno, cntry, blgetmg, imbgeco, imueclt, imwbcnt, gndr, agea, maritalb, eisced, polintr, hincfel, uemp5yr))

In your reports, please do the same for all data manipulation, such as subsetting or merging datasets - display r chunks, but hide the results.

Do not just use summary option

The summary option is well known to you from previous R classes. Let’s produce a summary table of all variables in our smaller dataset and a seperate one for variable age.

summary(round_8_final)

      idno              cntry              blgetmg         imbgeco      
 Min.   :        1   Length:44387       Min.   :1.000   Min.   : 0.000  
 1st Qu.:     1208   Class :character   1st Qu.:2.000   1st Qu.: 3.000  
 Median :     2589   Mode  :character   Median :2.000   Median : 5.000  
 Mean   : 31545782                      Mean   :1.935   Mean   : 5.006  
 3rd Qu.:    11058                      3rd Qu.:2.000   3rd Qu.: 7.000  
 Max.   :551603139                      Max.   :2.000   Max.   :10.000  
    imueclt          imwbcnt            gndr            agea       
 Min.   : 0.000   Min.   : 0.000   Min.   :1.000   Min.   : 15.00  
 1st Qu.: 4.000   1st Qu.: 3.000   1st Qu.:1.000   1st Qu.: 34.00  
 Median : 5.000   Median : 5.000   Median :2.000   Median : 49.00  
 Mean   : 5.368   Mean   : 4.904   Mean   :1.526   Mean   : 49.14  
 3rd Qu.: 7.000   3rd Qu.: 7.000   3rd Qu.:2.000   3rd Qu.: 64.00  
 Max.   :10.000   Max.   :10.000   Max.   :2.000   Max.   :100.00  
    maritalb         eisced          polintr         hincfel     
 Min.   :1.000   Min.   : 1.000   Min.   :1.000   Min.   :1.000  
 1st Qu.:1.000   1st Qu.: 2.000   1st Qu.:2.000   1st Qu.:1.000  
 Median :2.000   Median : 4.000   Median :3.000   Median :2.000  
 Mean   :3.153   Mean   : 4.109   Mean   :2.587   Mean   :1.948  
 3rd Qu.:6.000   3rd Qu.: 5.000   3rd Qu.:3.000   3rd Qu.:2.000  
 Max.   :6.000   Max.   :55.000   Max.   :4.000   Max.   :4.000  
    uemp5yr     
 Min.   :1.00   
 1st Qu.:1.00   
 Median :2.00   
 Mean   :1.53   
 3rd Qu.:2.00   
 Max.   :2.00   
 [ reached getOption("max.print") -- omitted 1 row ]

summary(round_8_final$agea)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
  15.00   34.00   49.00   49.14   64.00  100.00     155 

Both outputs are ‘readable’, but not presented in a publishable format.

In next sections I overview a few functions that are useful in producing nice tables, which are easily pinted in HTML format in R Markdown.

Producing tables with xtable

First, let’s use xtable function which exports tables to LaTeX or HTML format.

You need to install this package first on your machine (go back to Technical notes section to see names of packages). xtable overview is here: https://cran.r-project.org/web/packages/xtable/vignettes/xtableGallery.pdf

xtable needs to be supported with print function, which helps in ‘translating’ the output into HTML code associated with xtable object. See more here

I use this function to display first 10 rows of my data for all variables in the dataset. My r chunk is displayed first, then the result - I made them both visisble.

library("xtable") 
xtable_10rows <- xtable(round_8_final[1:10, ], digits = 0)
print(xtable_10rows, type = "html")

<!-- html table generated in R 3.6.3 by xtable 1.8-4 package -->
<!-- Fri May 01 15:12:50 2020 -->
<table border=1>
<tr> <th>  </th> <th> idno </th> <th> cntry </th> <th> blgetmg </th> <th> imbgeco </th> <th> imueclt </th> <th> imwbcnt </th> <th> gndr </th> <th> agea </th> <th> maritalb </th> <th> eisced </th> <th> polintr </th> <th> hincfel </th> <th> uemp5yr </th>  </tr>
  <tr> <td align="right"> 1 </td> <td align="right"> 1 </td> <td> AT </td> <td align="right"> 2 </td> <td align="right"> 10 </td> <td align="right"> 10 </td> <td align="right"> 10 </td> <td align="right"> 2 </td> <td align="right"> 34 </td> <td align="right"> 6 </td> <td align="right"> 7 </td> <td align="right"> 1 </td> <td align="right"> 2 </td> <td align="right"> 1 </td> </tr>
  <tr> <td align="right"> 2 </td> <td align="right"> 2 </td> <td> AT </td> <td align="right"> 2 </td> <td align="right"> 8 </td> <td align="right"> 8 </td> <td align="right"> 5 </td> <td align="right"> 1 </td> <td align="right"> 52 </td> <td align="right"> 1 </td> <td align="right"> 4 </td> <td align="right"> 1 </td> <td align="right"> 2 </td> <td align="right"> 2 </td> </tr>
  <tr> <td align="right"> 3 </td> <td align="right"> 4 </td> <td> AT </td> <td align="right"> 2 </td> <td align="right"> 5 </td> <td align="right"> 5 </td> <td align="right"> 4 </td> <td align="right"> 2 </td> <td align="right"> 68 </td> <td align="right"> 6 </td> <td align="right"> 3 </td> <td align="right"> 3 </td> <td align="right"> 2 </td> <td align="right"> 2 </td> </tr>
  <tr> <td align="right"> 4 </td> <td align="right"> 6 </td> <td> AT </td> <td align="right"> 2 </td> <td align="right"> 7 </td> <td align="right"> 7 </td> <td align="right"> 10 </td> <td align="right"> 1 </td> <td align="right"> 54 </td> <td align="right"> 4 </td> <td align="right"> 3 </td> <td align="right"> 2 </td> <td align="right"> 2 </td> <td align="right">  </td> </tr>
  <tr> <td align="right"> 5 </td> <td align="right"> 10 </td> <td> AT </td> <td align="right"> 1 </td> <td align="right"> 2 </td> <td align="right"> 5 </td> <td align="right"> 5 </td> <td align="right"> 2 </td> <td align="right"> 20 </td> <td align="right"> 1 </td> <td align="right"> 3 </td> <td align="right"> 3 </td> <td align="right"> 2 </td> <td align="right">  </td> </tr>
  <tr> <td align="right"> 6 </td> <td align="right"> 11 </td> <td> AT </td> <td align="right"> 2 </td> <td align="right"> 8 </td> <td align="right"> 6 </td> <td align="right"> 8 </td> <td align="right"> 2 </td> <td align="right"> 65 </td> <td align="right"> 4 </td> <td align="right"> 4 </td> <td align="right"> 2 </td> <td align="right"> 1 </td> <td align="right">  </td> </tr>
  <tr> <td align="right"> 7 </td> <td align="right"> 12 </td> <td> AT </td> <td align="right"> 1 </td> <td align="right"> 6 </td> <td align="right"> 5 </td> <td align="right"> 3 </td> <td align="right"> 2 </td> <td align="right"> 52 </td> <td align="right"> 1 </td> <td align="right"> 2 </td> <td align="right"> 3 </td> <td align="right"> 3 </td> <td align="right">  </td> </tr>
  <tr> <td align="right"> 8 </td> <td align="right"> 13 </td> <td> AT </td> <td align="right"> 2 </td> <td align="right"> 5 </td> <td align="right"> 4 </td> <td align="right"> 3 </td> <td align="right"> 2 </td> <td align="right"> 44 </td> <td align="right"> 1 </td> <td align="right"> 5 </td> <td align="right"> 3 </td> <td align="right"> 2 </td> <td align="right">  </td> </tr>
  <tr> <td align="right"> 9 </td> <td align="right"> 14 </td> <td> AT </td> <td align="right"> 2 </td> <td align="right"> 9 </td> <td align="right"> 10 </td> <td align="right"> 9 </td> <td align="right"> 2 </td> <td align="right"> 22 </td> <td align="right"> 6 </td> <td align="right"> 3 </td> <td align="right"> 4 </td> <td align="right"> 1 </td> <td align="right">  </td> </tr>
  <tr> <td align="right"> 10 </td> <td align="right"> 15 </td> <td> AT </td> <td align="right"> 2 </td> <td align="right"> 7 </td> <td align="right"> 9 </td> <td align="right"> 9 </td> <td align="right"> 2 </td> <td align="right"> 41 </td> <td align="right"> 4 </td> <td align="right"> 3 </td> <td align="right"> 2 </td> <td align="right"> 2 </td> <td align="right"> 1 </td> </tr>
   </table>

Ups - this does not look so well. You have to specify results='asis' in the r chunk. The deafault is results='markup' which writes the resutls in the raw html form (which is read by LaTeX). The table looks like that when you specify results='asis':

library("xtable") 
xtable_10rows <- xtable(round_8_final[1:10, ], digits = 0)
print(xtable_10rows, type = "html")

	idno	cntry	blgetmg	imbgeco	imueclt	imwbcnt	gndr	agea	maritalb	eisced	polintr	hincfel	uemp5yr
1	1	AT	2	10	10	10	2	34	6	7	1	2	1
2	2	AT	2	8	8	5	1	52	1	4	1	2	2
3	4	AT	2	5	5	4	2	68	6	3	3	2	2
4	6	AT	2	7	7	10	1	54	4	3	2	2
5	10	AT	1	2	5	5	2	20	1	3	3	2
6	11	AT	2	8	6	8	2	65	4	4	2	1
7	12	AT	1	6	5	3	2	52	1	2	3	3
8	13	AT	2	5	4	3	2	44	1	5	3	2
9	14	AT	2	9	10	9	2	22	6	3	4	1
10	15	AT	2	7	9	9	2	41	4	3	2	2	1

Let’s now hide the code (option echo=FALSE in r chunk) and add table name before the output. Option caption in xtable allows adding title, but it apears below the table. You can however add one more option to print - caption.placement="top". Finally, you can use option align in xtable to indicate whether you want text in cells to be placed in the centre (“c”), on the right (“r”) or on the left (“l”). This has to be specified for all columns and one additional column at the begining (so number of columns + 1 times).

Table 1. Overview of 10 first rows in data

	idno	cntry	blgetmg	imbgeco	imueclt	imwbcnt	gndr	agea	maritalb	eisced	polintr	hincfel	uemp5yr
1	1	AT	2	10	10	10	2	34	6	7	1	2	1
2	2	AT	2	8	8	5	1	52	1	4	1	2	2
3	4	AT	2	5	5	4	2	68	6	3	3	2	2
4	6	AT	2	7	7	10	1	54	4	3	2	2
5	10	AT	1	2	5	5	2	20	1	3	3	2
6	11	AT	2	8	6	8	2	65	4	4	2	1
7	12	AT	1	6	5	3	2	52	1	2	3	3
8	13	AT	2	5	4	3	2	44	1	5	3	2
9	14	AT	2	9	10	9	2	22	6	3	4	1
10	15	AT	2	7	9	9	2	41	4	3	2	2	1

Data overview with kable

It turns out that kable function poduces much better looking tables in the HTML format (xtable is preffered for PDF/LaTeX formats).

This section of R Markdown book provides more examples on how to use kable: https://bookdown.org/yihui/rmarkdown-cookbook/kable.html

Let’s display the same first 10 rows of our data, as previously, but now using kable function. Table title is more straighforward option here.

library("knitr")
kable(round_8_final[1:10, ], caption = "Table 1. Overview of 10 first rows in data") 

Table 1. Overview of 10 first rows in data

idno	cntry	blgetmg	imbgeco	imueclt	imwbcnt	gndr	agea	maritalb	eisced	polintr	hincfel	uemp5yr
1	AT	2	10	10	10	2	34	6	7	1	2	1
2	AT	2	8	8	5	1	52	1	4	1	2	2
4	AT	2	5	5	4	2	68	6	3	3	2	2
6	AT	2	7	7	10	1	54	4	3	2	2	NA
10	AT	1	2	5	5	2	20	1	3	3	2	NA
11	AT	2	8	6	8	2	65	4	4	2	1	NA
12	AT	1	6	5	3	2	52	1	2	3	3	NA
13	AT	2	5	4	3	2	44	1	5	3	2	NA
14	AT	2	9	10	9	2	22	6	3	4	1	NA
15	AT	2	7	9	9	2	41	4	3	2	2	1

Summary statistics with summarytools package

Before diving into any more complex data modelling, you should always explore frequency distribution of your key variables and compute desriptive statistics in order to understand what is happening in the data better.

summarytools package offers a wide range of functions to do so:

• package overview is here https://cran.r-project.org/web/packages/summarytools/summarytools.pdf
• R Markdown recommendations: https://cran.r-project.org/web/packages/summarytools/vignettes/Recommendations-rmarkdown.html

Descriptive statistics for all variables in your data - descr

Function descr produces a table with all descritpive statistics for all variables. In most functions from summarytools package you have to specify style = "rmarkdown" (but see below), so it produces nice looking tables in this format.

library("summarytools")
descr(round_8_final, style = 'rmarkdown')

Descriptive Statistics

round_8_final

N: 44387

Table continues below

	agea	blgetmg	eisced	gndr	hincfel	idno
Mean	49.14	1.94	4.11	1.53	1.95	31545782.19
Std.Dev	18.61	0.25	2.93	0.50	0.83	115541717.82
Min	15.00	1.00	1.00	1.00	1.00	1.00
Q1	34.00	2.00	2.00	1.00	1.00	1208.00
Median	49.00	2.00	4.00	2.00	2.00	2589.00
Q3	64.00	2.00	5.00	2.00	2.00	11058.00
Max	100.00	2.00	55.00	2.00	4.00	551603139.00
MAD	22.24	0.00	1.48	0.00	1.48	2640.51
IQR	30.00	0.00	3.00	1.00	1.00	9849.50
CV	0.38	0.13	0.71	0.33	0.43	3.66
Skewness	0.05	-3.54	10.44	-0.10	0.67	4.14
SE.Skewness	0.01	0.01	0.01	0.01	0.01	0.01
Kurtosis	-0.93	10.53	178.93	-1.99	-0.05	15.68
N.Valid	44232.00	43946.00	44258.00	44378.00	43863.00	44387.00
Pct.Valid	99.65	99.01	99.71	99.98	98.82	100.00

	imbgeco	imueclt	imwbcnt	maritalb	polintr	uemp5yr
Mean	5.01	5.37	4.90	3.15	2.59	1.53
Std.Dev	2.52	2.64	2.40	2.26	0.92	0.50
Min	0.00	0.00	0.00	1.00	1.00	1.00
Q1	3.00	4.00	3.00	1.00	2.00	1.00
Median	5.00	5.00	5.00	2.00	3.00	2.00
Q3	7.00	7.00	7.00	6.00	3.00	2.00
Max	10.00	10.00	10.00	6.00	4.00	2.00
MAD	2.97	2.97	2.97	1.48	1.48	0.00
IQR	4.00	3.00	4.00	5.00	1.00	1.00
CV	0.50	0.49	0.49	0.72	0.36	0.33
Skewness	-0.24	-0.28	-0.16	0.22	-0.03	-0.13
SE.Skewness	0.01	0.01	0.01	0.01	0.01	0.02
Kurtosis	-0.51	-0.60	-0.29	-1.80	-0.86	-1.98
N.Valid	42825.00	42984.00	42825.00	43509.00	44290.00	12406.00
Pct.Valid	96.48	96.84	96.48	98.02	99.78	27.95

As we do not want to display so much statistical informaiton we can constrain stats options to more basic descriptive statistics: mean, standard deviation, minimum and maximum value. I specify to swap columns with rows (transpose), hide headings and I add title with caption.

descr(round_8_final, stats = c("mean", "sd", "min", "max"), transpose = TRUE, headings = FALSE, caption="Table 1. Descriptive statistics for key variables")

Table 1. Descriptive statistics for key variables

	Mean	Std.Dev	Min	Max
agea	49.14	18.61	15.00	100.00
blgetmg	1.94	0.25	1.00	2.00
eisced	4.11	2.93	1.00	55.00
gndr	1.53	0.50	1.00	2.00
hincfel	1.95	0.83	1.00	4.00
idno	31545782.19	115541717.82	1.00	551603139.00
imbgeco	5.01	2.52	0.00	10.00
imueclt	5.37	2.64	0.00	10.00
imwbcnt	4.90	2.40	0.00	10.00
maritalb	3.15	2.26	1.00	6.00
polintr	2.59	0.92	1.00	4.00
uemp5yr	1.53	0.50	1.00	2.00

Frequency distribution - freq

If you want to display frequency distribution for one variable, use fuction freq

freq(round_8_final$gndr, style = "rmarkdown")

Frequencies

round_8_final$gndr

Label: Gender

	Freq	% Valid	% Valid Cum.	% Total	% Total Cum.
1	21027	47.38	47.38	47.37	47.37
2	23351	52.62	100.00	52.61	99.98
<NA>	9			0.02	100.00
Total	44387	100.00	100.00	100.00	100.00

Let’s add table title with caption option, and in order to remove the additional information before the table (e.g. round_8_final$gndr), specify option headings = FALSE in freq. I will add it to all remaining examples below.

freq(round_8_final$gndr, style = "rmarkdown", headings = FALSE, caption="Table 1. Sample distribution by gender, ESS 2016")

Table 1. Sample distribution by gender, ESS 2016

	Freq	% Valid	% Valid Cum.	% Total	% Total Cum.
1	21027	47.38	47.38	47.37	47.37
2	23351	52.62	100.00	52.61	99.98
<NA>	9			0.02	100.00
Total	44387	100.00	100.00	100.00	100.00

Cross-tabulate two variables - ctable

When you want to display proportions by groups use ctable function. You can specify pecentages to be calculate for rows (prop="r") or columns (prop="c").

ctable(round_8_final$gndr, round_8_final$hincfel, style = "rmarkdown", prop = "r", headings = FALSE, caption="Table 1. Subjective income by gender in ESS 2016")

Table 1. Subjective income by gender in ESS 2016

	hincfel	1	2	3	4	<NA>	Total
gndr
1		7101 (33.8%)	9807 (46.6%)	2906 (13.8%)	973 (4.6%)	240 ( 1.1%)	21027 (100.0%)
2		6922 (29.6%)	10684 (45.8%)	4039 (17.3%)	1427 (6.1%)	279 ( 1.2%)	23351 (100.0%)
<NA>		3 (33.3%)	1 (11.1%)	0 ( 0.0%)	0 (0.0%)	5 (55.6%)	9 (100.0%)
Total		14026 (31.6%)	20492 (46.2%)	6945 (15.6%)	2400 (5.4%)	524 ( 1.2%)	44387 (100.0%)

Data frame summary - dfSummary

This function prodces summary of a data frame consisting of: variable names, labels if any, factor levels, frequencies and/or numerical summary statistics, and valid/missing observation counts. plain.ascii = FALSE makes the graphs look better and style = 'grid' is recommended.

dfSummary(round_8_final, plain.ascii = FALSE, style = 'grid', graph.magnif = 0.75, valid.col = FALSE, tmp.img.dir = "/tmp", headings = FALSE, caption="Table 1. Summary of data frame")

Table 1. Summary of data frame

No	Variable	Label	Stats / Values	Freqs (% of Valid)	Graph	Missing
1	idno [numeric]		Mean (sd) : 31545782.2 (115541717.8) min < med < max: 1 < 2589 < 551603139 IQR (CV) : 9849.5 (3.7)	18328 distinct values		0 (0%)
2	cntry [character]		1. DE 2. IE 3. IT 4. IL 5. RU 6. CZ 7. LT 8. FR 9. EE 10. AT [ 13 others ]	2852 ( 6.4%) 2757 ( 6.2%) 2626 ( 5.9%) 2557 ( 5.8%) 2430 ( 5.5%) 2269 ( 5.1%) 2122 ( 4.8%) 2070 ( 4.7%) 2019 ( 4.5%) 2010 ( 4.5%) 20675 (46.6%)		0 (0%)
3	blgetmg [haven_labelled]	Belong to minority ethnic group in country	Min : 1 Mean : 1.9 Max : 2	1 : 2843 ( 6.5%) 2 : 41103 (93.5%)		441 (0.99%)
4	imbgeco [haven_labelled]	Immigration bad or good for country’s economy	Mean (sd) : 5 (2.5) min < med < max: 0 < 5 < 10 IQR (CV) : 4 (0.5)	11 distinct values		1562 (3.52%)
5	imueclt [haven_labelled]	Country’s cultural life undermined or enriched by immigrants	Mean (sd) : 5.4 (2.6) min < med < max: 0 < 5 < 10 IQR (CV) : 3 (0.5)	11 distinct values		1403 (3.16%)
6	imwbcnt [haven_labelled]	Immigrants make country worse or better place to live	Mean (sd) : 4.9 (2.4) min < med < max: 0 < 5 < 10 IQR (CV) : 4 (0.5)	11 distinct values		1562 (3.52%)
7	gndr [haven_labelled]	Gender	Min : 1 Mean : 1.5 Max : 2	1 : 21027 (47.4%) 2 : 23351 (52.6%)		9 (0.02%)
8	agea [haven_labelled]	Age of respondent, calculated	Mean (sd) : 49.1 (18.6) min < med < max: 15 < 49 < 100 IQR (CV) : 30 (0.4)	86 distinct values		155 (0.35%)
9	maritalb [haven_labelled]	Legal marital status, post coded	Mean (sd) : 3.2 (2.3) min < med < max: 1 < 2 < 6 IQR (CV) : 5 (0.7)	1 : 21711 (49.9%) 2 : 443 ( 1.0%) 3 : 648 ( 1.5%) 4 : 3912 ( 9.0%) 5 : 3756 ( 8.6%) 6 : 13039 (30.0%)		878 (1.98%)
10	eisced [haven_labelled]	Highest level of education, ES - ISCED	Mean (sd) : 4.1 (2.9) min < med < max: 1 < 4 < 55 IQR (CV) : 3 (0.7)	1 : 3861 ( 8.7%) 2 : 7388 (16.7%) 3 : 7153 (16.2%) 4 : 8720 (19.7%) 5 : 6275 (14.2%) 6 : 4760 (10.8%) 7 : 6013 (13.6%) 55 : 88 ( 0.2%)		129 (0.29%)
11	polintr [haven_labelled]	How interested in politics	Mean (sd) : 2.6 (0.9) min < med < max: 1 < 3 < 4 IQR (CV) : 1 (0.4)	1 : 5415 (12.2%) 2 : 15539 (35.1%) 3 : 15248 (34.4%) 4 : 8088 (18.3%)		97 (0.22%)
12	hincfel [haven_labelled]	Feeling about household’s income nowadays	Mean (sd) : 1.9 (0.8) min < med < max: 1 < 2 < 4 IQR (CV) : 1 (0.4)	1 : 14026 (32.0%) 2 : 20492 (46.7%) 3 : 6945 (15.8%) 4 : 2400 ( 5.5%)		524 (1.18%)
13	uemp5yr [haven_labelled]	Any period of unemployment and work seeking within last 5 years	Min : 1 Mean : 1.5 Max : 2	1 : 5794 (46.7%) 2 : 6612 (53.3%)		31981 (72.05%)

This looks great, but might take too much space.

As this produces large table in the middle of your report, you might either display it at the end in Appendix, or constrain height of the table, by embedding dfSummary in print function.

In the table below I also excluded the column with numbers (varnumbers = FALSE), and constrained the width of each column (col.widths = 25, ...). It has to be specified for each remaining column.

print(dfSummary(round_8_final, plain.ascii = FALSE, style = 'grid', graph.magnif = 0.85, valid.col = FALSE, tmp.img.dir = "/tmp", varnumbers = FALSE, col.widths = c(35, 35, 35, 35, 35, 35), headings = FALSE, caption="Table 1. Summary of data frame"), 
      max.tbl.height = 400, method = "render")

Variable	Label	Stats / Values	Freqs (% of Valid)	Graph	Missing
idno [numeric]		Mean (sd) : 31545782.2 (115541717.8) min < med < max: 1 < 2589 < 551603139 IQR (CV) : 9849.5 (3.7)	18328 distinct values		0 (0%)
cntry [character]		1. DE 2. IE 3. IT 4. IL 5. RU 6. CZ 7. LT 8. FR 9. EE 10. AT [ 13 others ]	2852 ( 6.4% ) 2757 ( 6.2% ) 2626 ( 5.9% ) 2557 ( 5.8% ) 2430 ( 5.5% ) 2269 ( 5.1% ) 2122 ( 4.8% ) 2070 ( 4.7% ) 2019 ( 4.5% ) 2010 ( 4.5% ) 20675 ( 46.6% )		0 (0%)
blgetmg [haven_labelled]	Belong to minority ethnic group in country	Min : 1 Mean : 1.9 Max : 2	1 : 2843 ( 6.5% ) 2 : 41103 ( 93.5% )		441 (0.99%)
imbgeco [haven_labelled]	Immigration bad or good for country's economy	Mean (sd) : 5 (2.5) min < med < max: 0 < 5 < 10 IQR (CV) : 4 (0.5)	11 distinct values		1562 (3.52%)
imueclt [haven_labelled]	Country's cultural life undermined or enriched by immigrants	Mean (sd) : 5.4 (2.6) min < med < max: 0 < 5 < 10 IQR (CV) : 3 (0.5)	11 distinct values		1403 (3.16%)
imwbcnt [haven_labelled]	Immigrants make country worse or better place to live	Mean (sd) : 4.9 (2.4) min < med < max: 0 < 5 < 10 IQR (CV) : 4 (0.5)	11 distinct values		1562 (3.52%)
gndr [haven_labelled]	Gender	Min : 1 Mean : 1.5 Max : 2	1 : 21027 ( 47.4% ) 2 : 23351 ( 52.6% )		9 (0.02%)
agea [haven_labelled]	Age of respondent, calculated	Mean (sd) : 49.1 (18.6) min < med < max: 15 < 49 < 100 IQR (CV) : 30 (0.4)	86 distinct values		155 (0.35%)
maritalb [haven_labelled]	Legal marital status, post coded	Mean (sd) : 3.2 (2.3) min < med < max: 1 < 2 < 6 IQR (CV) : 5 (0.7)	1 : 21711 ( 49.9% ) 2 : 443 ( 1.0% ) 3 : 648 ( 1.5% ) 4 : 3912 ( 9.0% ) 5 : 3756 ( 8.6% ) 6 : 13039 ( 30.0% )		878 (1.98%)
eisced [haven_labelled]	Highest level of education, ES - ISCED	Mean (sd) : 4.1 (2.9) min < med < max: 1 < 4 < 55 IQR (CV) : 3 (0.7)	1 : 3861 ( 8.7% ) 2 : 7388 ( 16.7% ) 3 : 7153 ( 16.2% ) 4 : 8720 ( 19.7% ) 5 : 6275 ( 14.2% ) 6 : 4760 ( 10.8% ) 7 : 6013 ( 13.6% ) 55 : 88 ( 0.2% )		129 (0.29%)
polintr [haven_labelled]	How interested in politics	Mean (sd) : 2.6 (0.9) min < med < max: 1 < 3 < 4 IQR (CV) : 1 (0.4)	1 : 5415 ( 12.2% ) 2 : 15539 ( 35.1% ) 3 : 15248 ( 34.4% ) 4 : 8088 ( 18.3% )		97 (0.22%)
hincfel [haven_labelled]	Feeling about household's income nowadays	Mean (sd) : 1.9 (0.8) min < med < max: 1 < 2 < 4 IQR (CV) : 1 (0.4)	1 : 14026 ( 32.0% ) 2 : 20492 ( 46.7% ) 3 : 6945 ( 15.8% ) 4 : 2400 ( 5.5% )		524 (1.18%)
uemp5yr [haven_labelled]	Any period of unemployment and work seeking within last 5 years	Min : 1 Mean : 1.5 Max : 2	1 : 5794 ( 46.7% ) 2 : 6612 ( 53.3% )		31981 (72.05%)

Generated by summarytools 0.9.6 (R version 3.6.3)
2020-05-01

Regression results with kable

kable can be used to disply results of models. You have to first prepare a summary of the model.

model1 <- lm(imueclt ~ agea + gndr, data=round_8_final)
regression1 <- coef(summary(model1))
kable(regression1, caption = "Table 1. Linear regression model predicting attitudes to immigration")

Table 1. Linear regression model predicting attitudes to immigration

	Estimate	Std. Error	t value	Pr(>|t|)
(Intercept)	5.7416851	0.0520508	110.309299	0.0000000
agea	-0.0099040	0.0006862	-14.432407	0.0000000
gndr	0.0741516	0.0254428	2.914445	0.0035651

It does not dispaly automatically \(R^2\) statistic, so you would have to extract it with a code:

summary(model1)$r.squared

[1] 0.004975681

Let’s add this information now to the table.

model1 <- lm(imueclt ~ agea + gndr, data=round_8_final)
regression1 <- coef(summary(model1))
kable(regression1, caption = "Table 1. Linear regression model predicting attitudes to immigration")

Table 1. Linear regression model predicting attitudes to immigration

	Estimate	Std. Error	t value	Pr(>|t|)
(Intercept)	5.7416851	0.0520508	110.309299	0.0000000
agea	-0.0099040	0.0006862	-14.432407	0.0000000
gndr	0.0741516	0.0254428	2.914445	0.0035651
\(R^2\)=0.005;	N=42842.

More kable options are discussed here: https://bookdown.org/yihui/rmarkdown-cookbook/kable.html

More advanced option and changing table style are here: https://cran.r-project.org/web/packages/kableExtra/vignettes/awesome_table_in_html.html#overview

Regression results with tab_model

The regression table prepared with kable does not look bad, but it can look even better with tab_model fucntion avaiable in sjPlot library. You can plot more than one model in the table and it automatically adds a few summary statistics on model fit.

library("sjPlot")
model2 <- lm(imueclt ~ agea + gndr, data=round_8_final)
model3 <- lm(imwbcnt ~ agea + gndr, data=round_8_final)
tab_model(model2, model3)

	Country’s cultural life undermined or enriched by immigrants			Immigrants make country worse or better place to live
Predictors	Estimates	CI	p	Estimates	CI	p
(Intercept)	5.74	5.64 – 5.84	<0.001	5.37	5.27 – 5.46	<0.001
Age of respondent, calculated	-0.01	-0.01 – -0.01	<0.001	-0.01	-0.01 – -0.01	<0.001
Gender	0.07	0.02 – 0.12	0.004	0.01	-0.03 – 0.06	0.567
Observations	42842			42679
R2 / R2 adjusted	0.005 / 0.005			0.006 / 0.006

This is a very good website which introduces more options, such as changing the default settings, removing/adding or defining own labels: https://cran.r-project.org/web/packages/sjPlot/vignettes/tab_model_estimates.html. You will notice that tab_model works with many types of regressions, including multilevel models.

Let’s add one more variable to the model above - hincfel, which measures relative deprivation as a ordinal variable (4 categories). I had to first tell R to treat it as a factor variable.

round_8_final$hincfel <- factor(round_8_final$hincfel)
model3 <- lm(imwbcnt ~ agea + gndr, data=round_8_final)
model4 <- lm(imwbcnt ~ agea + gndr + hincfel, data=round_8_final)
tab_model(model3, model4)

	Immigrants make country worse or better place to live			Immigrants make country worse or better place to live
Predictors	Estimates	CI	p	Estimates	CI	p
(Intercept)	5.37	5.27 – 5.46	<0.001	5.86	5.76 – 5.95	<0.001
Age of respondent, calculated	-0.01	-0.01 – -0.01	<0.001	-0.01	-0.01 – -0.01	<0.001
Gender	0.01	-0.03 – 0.06	0.567	0.08	0.04 – 0.13	<0.001
hincfel: hincfel2				-0.75	-0.80 – -0.69	<0.001
hincfel: hincfel3				-1.36	-1.43 – -1.29	<0.001
hincfel: hincfel4				-1.82	-1.93 – -1.72	<0.001
Observations	42679			42254
R2 / R2 adjusted	0.006 / 0.006			0.056 / 0.056

Finally, I add show.reflvl = TRUE option, which adds rows to the table with referece categories for caegorical variables.

tab_model(model3, model4, 
  pred.labels = c("Intercept", "Age", "Female", "Ref. Living comfortably: Coping",
                  "Difficult", "Very difficult"),
  dv.labels = c("Model 1", "Model 2"),
  string.est = "Coeffcient",
  string.ci = "Conf. Int (95%)",
  string.p = "P-Value"
)

	Model 1			Model 2
Predictors	Coeffcient	Conf. Int (95%)	P-Value	Coeffcient	Conf. Int (95%)	P-Value
Intercept	5.37	5.27 – 5.46	<0.001	5.86	5.76 – 5.95	<0.001
Age	-0.01	-0.01 – -0.01	<0.001	-0.01	-0.01 – -0.01	<0.001
Female	0.01	-0.03 – 0.06	0.567	0.08	0.04 – 0.13	<0.001
Ref. Living comfortably: Coping				-0.75	-0.80 – -0.69	<0.001
Difficult				-1.36	-1.43 – -1.29	<0.001
Very difficult				-1.82	-1.93 – -1.72	<0.001
Observations	42679			42254
R2 / R2 adjusted	0.006 / 0.006			0.056 / 0.056

Same model but with stars to indicate the level of significance p.style = "a" and I also add title.

tab_model(model3, model4, 
  pred.labels = c("Intercept", "Age", "Female", "Ref. Living comfortably: </br> Coping",
                  "Difficult", "Very difficult"),
  dv.labels = c("Model 1", "Model 2"),
  string.est = "Coeffcient",
  string.ci = "Conf. Int (95%)",
  p.style = "a", title = "Table 1. Linear regression models prediting attitudes to immigration (Immigrants make country worse or better place to live (0-10))"
)

Table 1. Linear regression models prediting attitudes to immigration (Immigrants make country worse or better place to live (0-10))

	Model 1		Model 2
Predictors	Coeffcient	Conf. Int (95%)	Coeffcient	Conf. Int (95%)
Intercept	5.37 ***	5.27 – 5.46	5.86 ***	5.76 – 5.95
Age	-0.01 ***	-0.01 – -0.01	-0.01 ***	-0.01 – -0.01
Female	0.01	-0.03 – 0.06	0.08 ***	0.04 – 0.13
Ref. Living comfortably: Coping			-0.75 ***	-0.80 – -0.69
Difficult			-1.36 ***	-1.43 – -1.29
Very difficult			-1.82 ***	-1.93 – -1.72
Observations	42679		42254
R2 / R2 adjusted	0.006 / 0.006		0.056 / 0.056
p<0.05   ** p<0.01   *** p<0.001

As you can see in r chunk I add </br> HTML language command in the middle of the first subjective income label to split it into two rows. wrap.labels option might be handy in case of long labels. All tab_model options are listed in R documentation here: https://www.rdocumentation.org/packages/sjPlot/versions/2.8.3/topics/tab_model

Excercise

Using the same ESS 2016 data prepare a short report in R Markdown consisting of the following:

1. Subset ESS data for one country (you choose which one)
2. Prepare a summary table displaying frequency distribution of variable measuring interest in politics (polintr)
3. Crosstabulate this variable by gndr and prepare a table with row percentages

• visualise this (see examples in my next practical on figures)

4. For data from this country run two linear regressions which model the response for two variables measuring attitudes to immigration:

• imueclt - Country’s cultural life undermined or enriched by immigrants
• imbgeco - Immigration bad or good for country’s economy
• add agea, gndr, polintr as independent variables
• there might be a need to change polintr into a factor variable

5. Display the resutls of both models as one table using tab_model function

• make the table look transparent and nice

6. Organise these results in sections (using hashtags) and save as a fresh report (delete my notes)

• it is up to you which sections or r chunks you will display and which hide

When you finish coding the results, go to YAML section:

• change the theme in YAML section into rmdformats::readthedown (install first rmdformats package)
• update authorship and date there
• experiment with any other settings of the YAML section, e.g. use a different highlight option

Finally, publish your report on https://Rpubs.com (create an account) and send me a link to what you have done before next virtual lab on 15th May, Friday.

Enjoy exploring R Markdown!