Tidyverse–Create

The Data

I’ll be using the data from SARS 2003 Outbreak Complete Dataset from Kaggle. The raw github link is here

Loading Packages

library(tidyverse)

## ── Attaching packages ─────────────────────────────────────────────────────────────── tidyverse 1.3.0 ──

## ✓ ggplot2 3.3.0     ✓ purrr   0.3.3
## ✓ tibble  2.1.3     ✓ dplyr   0.8.5
## ✓ tidyr   1.0.2     ✓ stringr 1.4.0
## ✓ readr   1.3.1     ✓ forcats 0.4.0

## ── Conflicts ────────────────────────────────────────────────────────────────── tidyverse_conflicts() ──
## x dplyr::filter() masks stats::filter()
## x dplyr::lag()    masks stats::lag()

library(tidyr)

Loading the data

df=read.csv('https://raw.githubusercontent.com/metis-macys-66898/data_607_sp2020/master/sars_2003_complete_dataset_clean.csv', stringsAsFactors 
 = F)

head(df)

##         Date              Country Cumulative.number.of.case.s. Number.of.deaths
## 1 2003-03-17              Germany                            1                0
## 2 2003-03-17               Canada                            8                2
## 3 2003-03-17            Singapore                           20                0
## 4 2003-03-17 Hong Kong SAR, China                           95                1
## 5 2003-03-17          Switzerland                            2                0
## 6 2003-03-17             Thailand                            1                0
##   Number.recovered
## 1                0
## 2                0
## 3                0
## 4                0
## 5                0
## 6                0

str(df)

## 'data.frame':    2538 obs. of  5 variables:
##  $ Date                        : chr  "2003-03-17" "2003-03-17" "2003-03-17" "2003-03-17" ...
##  $ Country                     : chr  "Germany" "Canada" "Singapore" "Hong Kong SAR, China" ...
##  $ Cumulative.number.of.case.s.: int  1 8 20 95 2 1 40 2 8 0 ...
##  $ Number.of.deaths            : int  0 2 0 1 0 0 1 0 2 0 ...
##  $ Number.recovered            : int  0 0 0 0 0 0 0 0 0 0 ...

as.tibble(df)

## Warning: `as.tibble()` is deprecated, use `as_tibble()` (but mind the new semantics).
## This warning is displayed once per session.

## # A tibble: 2,538 x 5
##    Date     Country      Cumulative.number.of… Number.of.deaths Number.recovered
##    <chr>    <chr>                        <int>            <int>            <int>
##  1 2003-03… Germany                          1                0                0
##  2 2003-03… Canada                           8                2                0
##  3 2003-03… Singapore                       20                0                0
##  4 2003-03… Hong Kong S…                    95                1                0
##  5 2003-03… Switzerland                      2                0                0
##  6 2003-03… Thailand                         1                0                0
##  7 2003-03… Viet Nam                        40                1                0
##  8 2003-03… Germany                          2                0                0
##  9 2003-03… Canada                           8                2                0
## 10 2003-03… China                            0                0                0
## # … with 2,528 more rows

Using dplyr: rename to rename variables by name

Usage

rename(.data, …)

select(.data, …)

Examples

rename( ) can be used to rename columns, except all columns not mentioned are kept

df1 <- df %>% rename("Cumulative_number_of_cases" = "Cumulative.number.of.case.s.")
names(df1)

## [1] "Date"                       "Country"                   
## [3] "Cumulative_number_of_cases" "Number.of.deaths"          
## [5] "Number.recovered"

Select( ) can be used to rename columns, though all columns not mentioned are dropped

df %>% select("Cumulative_number_of_cases" = "Cumulative.number.of.case.s.") %>% head(5)

##   Cumulative_number_of_cases
## 1                          1
## 2                          8
## 3                         20
## 4                         95
## 5                          2

Need to make sure df1 is not tibble.

Using dplyr: top_frac to select top (or bottom) n rows (by value)

Usage

top_frac(x, n, wt)

top_n(x, n, wt)

Examples

Wanted to get the top 30% of the countries by day in terms of cumulative number of cases

# The funciton count will give us a count of records by day, denoted in n. 
# group_by (x) followed by %>% will give you the result grouping by x
df1 %>% group_by (Date) %>% count

## # A tibble: 96 x 2
## # Groups:   Date [96]
##    Date           n
##    <chr>      <int>
##  1 2003-03-17     7
##  2 2003-03-18    10
##  3 2003-03-19    12
##  4 2003-03-20    13
##  5 2003-03-21    15
##  6 2003-03-22    15
##  7 2003-03-24    15
##  8 2003-03-25    14
##  9 2003-03-26    14
## 10 2003-03-27    15
## # … with 86 more rows

.3 * n( ) will give you 30% of row count. Thus the following will give you the top 30% of the countries per date in terms of cumulative number of cases.

#  Assign the result to df2  
df2 <- df1 %>% group_by(Date ) %>% top_n(.3 * n(), Cumulative_number_of_cases)
df2

## # A tibble: 774 x 5
## # Groups:   Date [96]
##    Date     Country       Cumulative_number_o… Number.of.deaths Number.recovered
##    <chr>    <chr>                        <int>            <int>            <int>
##  1 2003-03… Hong Kong SA…                   95                1                0
##  2 2003-03… Viet Nam                        40                1                0
##  3 2003-03… Singapore                       23                0                0
##  4 2003-03… Hong Kong SA…                  123                1                0
##  5 2003-03… Viet Nam                        57                1                0
##  6 2003-03… Hong Kong SA…                  150                5                0
##  7 2003-03… Singapore                       31                0                0
##  8 2003-03… Viet Nam                        56                2                0
##  9 2003-03… Hong Kong SA…                  173                6                0
## 10 2003-03… Singapore                       34                0                0
## # … with 764 more rows

top_frac is basically a shorthand to eliminate the need to put in n() for the total row count in that group. Simply put in the fraction as a the parameter. See below.

# note that top_frac takes in .3 instead of .3 * n(). Assign the result to df2a
df2a <- df1 %>% group_by(Date ) %>% top_frac(.3, Cumulative_number_of_cases)
df2a

## # A tibble: 774 x 5
## # Groups:   Date [96]
##    Date     Country       Cumulative_number_o… Number.of.deaths Number.recovered
##    <chr>    <chr>                        <int>            <int>            <int>
##  1 2003-03… Hong Kong SA…                   95                1                0
##  2 2003-03… Viet Nam                        40                1                0
##  3 2003-03… Singapore                       23                0                0
##  4 2003-03… Hong Kong SA…                  123                1                0
##  5 2003-03… Viet Nam                        57                1                0
##  6 2003-03… Hong Kong SA…                  150                5                0
##  7 2003-03… Singapore                       31                0                0
##  8 2003-03… Viet Nam                        56                2                0
##  9 2003-03… Hong Kong SA…                  173                6                0
## 10 2003-03… Singapore                       34                0                0
## # … with 764 more rows

# checking if df2a is the same as df2
all.equal(df2, df2a)

## [1] TRUE

Use dplyr::tally or dplyr::count to do counts like n( )

Besides summarize/summarise, what’re some of the functions that one can use to tally counts?

Usage

tally(x, wt = NULL, sort = FALSE, name = “n”)

count(x, …, wt = NULL, sort = FALSE, name = “n”, .drop = group_by_drop_default(x))

tally the record/obs count by country

# an automatic variable n is created to hold the results of tally
df2 %>% group_by(Country) %>% tally

## # A tibble: 12 x 2
##    Country                  n
##    <chr>                <int>
##  1 Canada                  82
##  2 China                   89
##  3 Germany                 53
##  4 Hong Kong SAR, China    96
##  5 Italy                   29
##  6 Mongolia                22
##  7 Philippines             52
##  8 Singapore               95
##  9 Taiwan, China           71
## 10 Thailand                 9
## 11 United States           89
## 12 Viet Nam                87

Count the number of record/obs by country

# It can be further simplified to using just count. count (x) means counting the number of records grouping by x. 
df2 %>% count (Country)

## # A tibble: 773 x 3
## # Groups:   Date [96]
##    Date       Country                  n
##    <chr>      <chr>                <int>
##  1 2003-03-17 Hong Kong SAR, China     1
##  2 2003-03-17 Viet Nam                 1
##  3 2003-03-18 Hong Kong SAR, China     1
##  4 2003-03-18 Singapore                1
##  5 2003-03-18 Viet Nam                 1
##  6 2003-03-19 Hong Kong SAR, China     1
##  7 2003-03-19 Singapore                1
##  8 2003-03-19 Viet Nam                 1
##  9 2003-03-20 Hong Kong SAR, China     1
## 10 2003-03-20 Singapore                1
## # … with 763 more rows

Tidyverse–Create

Dennis Pong

3/29/2020

The Data

Loading Packages

Loading the data

Using dplyr: rename to rename variables by name

Usage

Examples

rename( ) can be used to rename columns, except all columns not mentioned are kept

Select( ) can be used to rename columns, though all columns not mentioned are dropped

Need to make sure df1 is not tibble.

Using dplyr: top_frac to select top (or bottom) n rows (by value)

Usage

Examples

Wanted to get the top 30% of the countries by day in terms of cumulative number of cases

.3 * n( ) will give you 30% of row count. Thus the following will give you the top 30% of the countries per date in terms of cumulative number of cases.

top_frac is basically a shorthand to eliminate the need to put in n() for the total row count in that group. Simply put in the fraction as a the parameter. See below.

Use dplyr::tally or dplyr::count to do counts like n( )

Besides summarize/summarise, what’re some of the functions that one can use to tally counts?

Usage

tally the record/obs count by country

Count the number of record/obs by country