STAT545B Assignment B2- Aditi Nagaraj Nallan

Aditi Nagaraj Nallan 11/17/2021

Overview:

This github repository contains two components encompassing Assignments- B1 and B2 in the UBC graduate course STAT 545. The first component is a folder called Assignment_B1 which contains a function, its description and usage. This is included within a package called SummaryStat which is the second component of this repository.

A short description of the function along with its usage and test cases are provided in the path- Assignment_B1/Assignment_B1.md which was submitted as part of Assignment B1.

For Assignment B2, the aforementioned function was wrapped in an R package. A description of the package, installation instructions and example usage is provided below as part of this README file.

Package description:

Package name: SummaryStat

Package creator: Aditi Nagaraj Nallan

Package version: 0.1.0

SummaryStat is a package made by Aditi Nagaraj Nallan for the STAT545B graduate course at The University of British Columbia. This package was made in R using the devtools package and is mainly used for calculating various summary statistics across different numerical and categorical variables from input datasets. The main goal of this package is to make exploratory data analysis efficient and easy for users.

Currently, this package contains only one function called dplyr_bundle.

This function bundles together two dplyr functions- group_by and summarise to calculate the summary statistics such as mean and range for a numerical variable grouped by a categorical variable (both of which are user choice) from an input dataset. The input dataset can either be one of the publicly available datasets from packages such as gapminder, datateachr and palmerpenguins or any user-defined input dataset.

Further developments to the package will include addition of a function that counts the observations across specified categorical variables and another function that plots the various summary statistics.

Installation and example usage is shown below.

Installation:

Since the package is currently under development, only the development version from github is available. Once the package is submitted to CRAN a released version with updated installation instructions will be available.

The development version from GitHub can be installed as follows:

# install.packages("devtools")
devtools::install_github("stat545ubc-2021/SummaryStat")

## Downloading GitHub repo stat545ubc-2021/SummaryStat@HEAD

## 
##      checking for file ‘/tmp/RtmpDxQq7n/remotes362d71283e37/stat545ubc-2021-SummaryStat-6ea5417/DESCRIPTION’ ...  ✓  checking for file ‘/tmp/RtmpDxQq7n/remotes362d71283e37/stat545ubc-2021-SummaryStat-6ea5417/DESCRIPTION’
##   ─  preparing ‘SummaryStat’:
##      checking DESCRIPTION meta-information ...  ✓  checking DESCRIPTION meta-information
##   ─  checking for LF line-endings in source and make files and shell scripts
##   ─  checking for empty or unneeded directories
##   ─  building ‘SummaryStat_0.1.0.tar.gz’
##      
## 

## Installing package into '/tmp/Rtmpa9Dsod/temp_libpath332559bdbad9'
## (as 'lib' is unspecified)

Example Usage:

As specified in the package description, the dplyr_bundle function from the package can be used for publicly available datasets and user input datasets as well. The next few code chunks show instances of using the function for both of these cases.

# Before running the function load the package.
library(SummaryStat)

1. Publicly available datasets

There are many publicly available datasets within packages such as gapminder, palmerpenguins and datateachr. One of these datasets that I will show here is the penguins dataset from the palmerpenguins package. The following example uses the dplyr_bundle function to calculate mean and range of the flipper length of different penguin species. Here the categorical variable would be species and numerical variable would be flipper_length_mm.

SummaryStat::dplyr_bundle(palmerpenguins::penguins, group = species, summary = flipper_length_mm)

## # A tibble: 3 × 4
##   species   mean_flipper_length_mm min_flipper_length_mm max_flipper_length_mm
##   <fct>                      <dbl>                 <int>                 <int>
## 1 Adelie                      190.                   172                   210
## 2 Chinstrap                   196.                   178                   212
## 3 Gentoo                      217.                   203                   231

2. User input datasets

These datasets can be anything that the user is currently analyzing or will analyze in the future. An example dataset to test the function is provided along with this package and is located in the data-raw folder. The file is called Abundance.csv

data = read.csv("https://raw.githubusercontent.com/stat545ubc-2021/SummaryStat/main/data-raw/Abundance.csv", header = TRUE, sep = ",")
SummaryStat::dplyr_bundle(data, lineage, TPM)

## # A tibble: 4 × 4
##   lineage             mean_TPM min_TPM max_TPM
##   <fct>                  <dbl>   <dbl>   <dbl>
## 1 Bacteria                6.26    1.02    19.2
## 2 Bacteroidetes          72.4    41.1    133. 
## 3 Clostridiales          32.5     2.47   140. 
## 4 Deltaproteobacteria    14.7     1.45    53.3

To use your own dataset as shown above, read the file from the appropriate location on your system and store it. Load the package SummaryStat and then call the function dplyr_bundle to compute summary statistics for the stored data.