You are on page 1of 8

JSS Journal of Statistical Software

MMMMMM YYYY, Volume VV, Issue II. doi: 10.18637/jss.v000.i00

SWANSTAT: A User-Friendly Web Application for


Data Analysis using Shinydashboard

Rizal Bakri Bagus Sartono


STIEM Bongaya, Indonesia IPB University, Indonesia

Hazan Azhari Zainuddin Laode Ahmad Sabil


Swanstatistics, Indonesia Swanstatistics, Indonesia

Abstract
Technological advances gave birth to a million software to facilitate work in com-
plex data analysis. SWANSTAT is a user-friendly web application developed from the
R programming language that compiles from the best various packages and is built with
shinydashboard Package. The motivation to create SWANSTAT is to streamline the rou-
tine work flow of data analysis so that users unfamiliar with R can perform the analysis
interactively in a web browser with a high-ability server using a shiny server. SWANSTAT
is a new statistical software with various features including best visualization, the latest
statistical methods, help documents and tutorials, and can be accessed online using varios
types of browsers. This research will be continuous for the development of the application
by enriching the latest statistical method, as well as the improvement of the quality of
features for data science needs.

Keywords: SWANSTAT, Shinydashboard, Statistical Methods, R Language, Shiny Server,


Data Science.

1. Introduction
Technological developments are more and more increasing which is marked by the birth of
various sophisticated software as an aids to facilitate people in doing their works. Nowadays,
both comercial and open-source statistical softwares haves been been developed to analyze
data that is so complex. A software which has been becoming popular and used by many
statisticians and data scientist across nations is R that was first discovered by Ross Ihaka
and Robert Gentleman in 1992 from the Statistics Department, University of Auckland, New
2 SWANSTAT: A User-Friendly Web Application for Data Analysis using Shinydashboard

Zealand (R Core Team 2017). The R software is open-source software so it collects thousands
of packages made by various contributors in the world and has high quality in processing
large data. The users of R may utilize R studio and shinydashboard package (Chang and
Ribeiro 2018) to develop easy and interactive application which may assists data analyst to
handle their analytical problem by avoiding to build program script by themselves.
Based on the authors‘ experience in conducting analytics trainings, many people face a big
challenge in writing R programing due to their lack of knowledge in the language and pro-
graming structure. It is easier for them to use a click-and-drag application and not worrying
about how the application handle the computation. However, it is not easy for them to find
a user friendly software which is free (open source) and consist of various kinds of statistical
analyses.
This research aims to design an application by utilizing the sophistication of R software to be
more user friendly, namely creating statistical tool interfaces from various existing packages.
Then the developed application can be accessed online using a highly capable server with the
help of a shiny server called SWANSTAT. In addition, the SWANSTAT application is also
equipped with a help document in the form such white papers and video tutorials to facilitate
users and to provide examples of the implementation.

2. Outline of Usage

2.1. How to acces this software ?


SWANSTAT has been developed with the aim of facilitating users who are not familiar to build
program script and setup the application on dekstop. The use of this application does not need
to be installation. This application has been stored in a highly capable computing server that
can be accessed online using shiny server (R Studio Team 2016). SWANSTAT can be accessed
using various types of browsers on the page https://apps.swanstatistics.com/swanstat.
The SWANSTAT user interface is depicted Figure 1 after you have successfully accessed the
URL.

2.2. Preparation and import of data


Data preparation is an importan initial planning before processing data using this application.
The file type supported by this application is data that has csv extension (comma delimeted)
from microsoft Excel. For the next development, this application will be equipped with various
types of files that can be imported such as the file type from SAS, MINITAB, SPSS, XLSX,
SQL, and others. The initial step in using this application is to port data. Import data can
be done by selecting the Import Data menu then searching for data through the Browse
button. Determine the parameters needed according to data structures such as headers and
separators. After that select Upload Data on the Select data Active drop down menu.
In adition SWANSTAT has also prepared some data driven taken from various package data
and displayed on the menu as an exercise. Figure 2 shows the page for data import stages.

2.3. Analysis and Execute Application


After the data is imported, the next step is to choose the right method based on the purpose
Journal of Statistical Software 3

Figure 1: Dashboard of SWANSTAT

3
4

Figure 2: Import Data

of the analysis. For example, we will use data driven, iris, for analyzing correlation between
quantitative varaibles. Then the first step is to choose the Correlation menu then select the
Bivariate submenu. Enter all quantitative variables that will be correlated in the Select Mul-
tiple Variable box. For this research we will use the Person correlation method. Visualization
of the correlation analysis can be displyed via the Plot Options tab menu. For correlation
analysis, the main package of SWAN-STAT is corrplot (Wei and Simko 2017). Figure 3 shows
the stages of using correlation analysis with SWANSTAT. To run the application, select the
RUN button then the analysis output can be seen in the Summary and Plot tab menu as
shown in Figure 4.
4 SWANSTAT: A User-Friendly Web Application for Data Analysis using Shinydashboard

1
4

Figure 3: Correlation Menu

Figure 4: Output of Correlation Analysis

3. Feature of SWANSTAT
Each application certainly has its own advantages. By combining various packages inside the
R, SWANSTAT is equipped with various features provided to facilitate users in operating
and analyzing data. The features provided by this application include Visualization, More
Statistics Methods, Help Document, and Save Output Report.

3.1. Visualization
Interesting visualization can help users easily provide interpretation of an analysis. This ap-
plication is equipped with visualization that combines various packages of Plots in R including
graphics (R Core Team 2017), ggplot2 (Wickham, Chang, Henry, Pedersen, Takahashi, Wilke,
Journal of Statistical Software 5

Figure 5: More Visualization of SWANSTAT

and Woo 2018) and plotly package (Sievert, Parmer, Hocking, Chamberlain, Ram, Corvellec,
and Despouy 2017). Figure 5 shows some graphic forms available in various statistical tools
in SWANSTAT.

3.2. More statistics method


The progress of statistical methods makes the development of this application continues.The
adition of statistical tools in this application continues to be developed based on the needs
and scope of the analytical methods contained in R. SWANSTAT has been equipped with
various statistical methods frome some basic packages commonly used. Basic statistical meth-
ods that have been made like Summary with fBasic package (Wuertz 2017), Tabulation,
Chi-Square analysis, Metode Succesive Interval (MSI), Slovin, various methods for
Normality test, Graphs (Pie Charts, Bat Charts, Box Plot, Scatter Plots, His-
togram), Compare Means (One Sample T Test, Independent T Test, Paired T
Test), ANOVA, Design of Experiment, Correlation (Pearson, Sperman, Kendall-
Tau dengan corrplot package (Wei and Simko 2017), Linear Regression, Logistics Re-
gression with caret package (Kuhn 7), Instrument Test (Validity, Reliability), Struc-
tural Model (Path Analysis, Structural Equation Model, Confirmatory Factor
Analysis Models) dengan semMediation (Moon 2017) and lavaan (Rosseel 2018) package,
Cluster (Hierarchical, Nonhierarchical) with cluster package (Maechler, Rousseeuw,
Struyf, and Hubert 2018), Time Series with forecast and tseries package, and other meth-
ods and packages that are still under development.

3.3. Document Help


Document help is designed to make it easier for users to use this application. Document
Guide that has been made in the form of articles and video tutorials. To access the document,
SWANSTAT has prepared the HELP button at each Analysis Menu. When pressing the but-
ton, this application points to the url https://swanstatistics.com/category/swanstat/.
6 SWANSTAT: A User-Friendly Web Application for Data Analysis using Shinydashboard

Figure 6: Save Output Box Dialog

Meanwhile the tutorial video about the step by step of the use of this application can be
accessed SWAN Statistics Academy channel in Youtube.

3.4. Save output report


One of the main features of this application is that the output from the analysis can be
saved. Output document format in the form of html and word using rmarkdown package
(Allaire, Xie, McPherson, Luraschi, Ushey, Atkins, Wickam, Chng, Chang, and Lannone
2018). SWANSTAT application analysis output is the same as the output in R Studio which
is designed in table form. To use it, click the SAVE button after analyzing the data. Figure
6 shows the display box when you want to save the output from the analysis. Select the
document format then Download.

4. Conclussion and future work


This research is the begining of the application development plan. We have tried to develop
statistical applications that can be accessed online. The features we have created include
import data with csv file extension, Visualization of various package graphics, Some basic
statistical methods, Document Help in the form of articles and video tutorials, etc. For the
future work is complementing other statistical methods and more interesting features. We the
development team are opening opportunities for programmers and statisticians to contribute
to the development of this software.

Acknowledgments
The main idea of the development of this application came from the publicâĂŹs anxiety over
the use of coding-based software. The team expressed many tahnks to the IPB Statistics
Lecturers who have supported the development of this application. Especially to Mr. Prof.
Dr. Asep Saefuddin, Prof. Dr. Khairil Anwar Notodiputro, Bapak Dr. Anang kurnia and Mr.
Dr. Bagus Sartono who has become our advisor. We also say thanks to all the contributor
packages that we have used in developing this application.
Journal of Statistical Software 7

References

Allaire J, Xie Y, McPherson J, Luraschi J, Ushey K, Atkins A, Wickam H, Chng J, Chang


W, Lannone R (2018). rmarkdown : Dynamic Dosuments for R. R package version 1.11,
URL https://CRAN.R-project.org/package=rmarkdown.

Chang W, Ribeiro BB (2018). shinydashboard : Create Dashboard with ’shiny’. R package


version 0.7.1, URL https://CRAN.R-project.org/package=shinydashboard.

Kuhn M (7). caret : Classification and Regression Training. R package version 6.0-80, URL
https://CRAN.R-project.org/package=caret.

Maechler M, Rousseeuw P, Struyf A, Hubert M (2018). cluster : "Finding Groups in Data":


Cluster Analysis Extended Rousseeuw et al. R package version 2.0.7-1, URL https://CRAN.
R-project.org/package=cluster.

Moon KW (2017). semMediation : Visualize the Mediation Effect for Structural Equa-
tion Modelling. R package version 0.1.1, URL https://github.com/cardiomoon/
semMediation.

R Core Team (2017). R: A Language and Environment for Statistical Computing. R Founda-
tion for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.

Rosseel Y (2018). lavaan : Latent Variable Analysis. R package version 0.6-2, URL https:
//CRAN.R-project.org/package=lavaan.

R Studio Team (2016). Shiny Server Open Source v1.4.0 Administrator’s Guide. R Studio,
Northern Ave, Boston. URL http://rstudio.github.io/shiny-server/os/latest/.

Sievert C, Parmer C, Hocking T, Chamberlain S, Ram K, Corvellec M, Despouy P (2017).


plotly : Create Interactive Web Graphics via ’plotly.js’. R package version 4.7.1, URL
https://CRAN.R-project.org/package=plotly.

Wei T, Simko V (2017). An Introduction to corrplot Package. R package version 0.84, URL
https://CRAN.R-project.org/package=corrplot.

Wickham H, Chang W, Henry L, Pedersen TL, Takahashi K, Wilke C, Woo K (2018). ggplot2
: Create Elegant Data Visualisations Using the Grammar of Graphics. R package version
3.0.0, URL https://CRAN.R-project.org/package=ggplot2.

Wuertz D (2017). fBasics : Rmetrics - Markets and Basic Statistics. R package version
3042.89, URL https://CRAN.R-project.org/package=fBasics.
8 SWANSTAT: A User-Friendly Web Application for Data Analysis using Shinydashboard

Affiliation:
Rizal Bakri
STIEM Bongaya
SWANSTATISTICS
IPB University
90223 Makassar, South Sulawesi, Indonesia
E-mail: rizal.bakri@stiem-bongaya.ac.id
URL: https://apps.swanstatistics.com/swanstat

Journal of Statistical Software http://www.jstatsoft.org/


published by the Foundation for Open Access Statistics http://www.foastat.org/
MMMMMM YYYY, Volume VV, Issue II Submitted: yyyy-mm-dd
doi:10.18637/jss.v000.i00 Accepted: yyyy-mm-dd

You might also like