README.md

In epicentre-msf/covidutils: Utility Functions For Epicentre COVID-19 Analysis

covidutils

Centralised utility functions for covid related projects at Epicentre.

Installation

Install covidutils from github with:

remotes::install_github("epicentre-msf/covidutils")

Development

To include a function in the package, add it to a script in the R/ directory.

Best practices

• Large functions should have their own .R script inside the R/ directory with an information name, get_ecdc.R for example
• Smaller utility functions can be grouped into a single script such as utils.R
• If you are using functions used from external packages inside your own function, they must be fully qualified i.e. dplyr::filter()
• All external packages used must be added to the imports section of the DESCRIPTION file. This can be done via usethis::use_package("name_of_package")
• All variables used within a function should be passed as arguments
• Document your functions using roxygen syntax
• When your changes are ready and documented:
  • run devtools::document() to build the appropriate documentation files
  • run devtools::check() to check the package for any errors or warnings
  • run devtools::install() to install or re-install the package on your system
  • consider adding a usage of example of any new functions in this README.Rmd below then knit the document to produce the .md format required for github
  • push the changes to github

If using RStudio, you can document, check and install the package with option in the ‘Build’ pane.

When developing new functions you can run devtools::load_all() at anytime load the current state of all function in the R/ directory.

Further R package development resources:

• Hadley Wickham’s R Packages book (free online)
• Karl Broman’s R package primer
• Hilary Parker’s R package from scratch tutorial

Function examples

library(tidyverse)
library(covidutils)

Import JHU CSSE data

df_covid <- get_owid_jhcsse()
glimpse(df_covid)
#> Rows: 84,398
#> Columns: 7
#> $ date      <date> 2020-02-24, 2020-02-25, 2020-02-26, 2020-02-27, 2020-02-28,…
#> $ continent <chr> "Asia", "Asia", "Asia", "Asia", "Asia", "Asia", "Asia", "Asi…
#> $ region    <chr> "Southern Asia", "Southern Asia", "Southern Asia", "Southern…
#> $ country   <chr> "Afghanistan", "Afghanistan", "Afghanistan", "Afghanistan", …
#> $ iso_a3    <chr> "AFG", "AFG", "AFG", "AFG", "AFG", "AFG", "AFG", "AFG", "AFG…
#> $ cases     <dbl> 1, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0, 0, 0, 1, 2, 1, 3, 1, 1, 2, …
#> $ deaths    <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, …

Compute 14 and 30 day case and death trends

df_trends <- get_country_summaries(df_covid)
glimpse(df_trends)
#> Rows: 190
#> Columns: 53
#> $ continent                    <chr> "Asia", "Europe", "Africa", "Europe", "Af…
#> $ region                       <chr> "Southern Asia", "Southern Europe", "Nort…
#> $ country                      <chr> "Afghanistan", "Albania", "Algeria", "And…
#> $ iso_a3                       <chr> "AFG", "ALB", "DZA", "AND", "AGO", "ATG",…
#> $ population                   <dbl> 38928341, 2877800, 43851043, 77265, 32866…
#> $ cases                        <dbl> 67743, 132244, 127646, 13671, 33338, 1258…
#> $ deaths                       <dbl> 2855, 2447, 3440, 127, 742, 42, 75588, 44…
#> $ cases_inc                    <dbl> 174.01975, 4595.31587, 291.09000, 17693.6…
#> $ deaths_inc                   <dbl> 7.3339884, 85.0302314, 7.8447393, 164.369…
#> $ date_start_14d               <date> 2021-05-11, 2021-05-11, 2021-05-11, 2021…
#> $ date_end_14d                 <date> 2021-05-24, 2021-05-24, 2021-05-24, 2021…
#> $ cases_14d                    <dbl> 4212, 462, 2819, 140, 3748, 21, 397014, 2…
#> $ trend_cases_14d              <chr> "Likely increasing", "Decreasing", "Incre…
#> $ trend_cases_coeff_14d        <dbl> 0.064730659, -0.063993259, 0.035337112, -…
#> $ trend_cases_coeff_lwr95_14d  <dbl> -0.0300305835, -0.0854030956, 0.004928549…
#> $ trend_cases_coeff_upr95_14d  <dbl> 0.159491902, -0.042583422, 0.065745676, 0…
#> $ trend_cases_preds_14d        <list> [<tbl_df[12 x 6]>], [<tbl_df[12 x 6]>], …
#> $ dt_cases_est_14d             <dbl> 10.708174, NA, 19.615275, NA, NA, NA, 23.…
#> $ dt_cases_lwr_14d             <dbl> 4.345971, NA, 10.542856, NA, NA, NA, 11.4…
#> $ dt_cases_upr_14d             <dbl> -23.08138, NA, 140.63919, NA, NA, NA, -90…
#> $ deaths_14d                   <dbl> 114, 29, 91, 0, 95, 10, 6659, 145, 0, 155…
#> $ cases_inc_14d                <dbl> 10.8375675, 16.8243617, 6.4469333, 219.66…
#> $ deaths_inc_14d               <dbl> 0.293355407, 1.056241202, 0.208126122, 0.…
#> $ trend_deaths_14d             <chr> "Increasing", NA, "Stable", NA, "Likely i…
#> $ trend_deaths_coeff_14d       <dbl> 0.0462455276, NA, -0.0007424017, NA, 0.06…
#> $ trend_deaths_coeff_lwr95_14d <dbl> 0.0123976954, NA, -0.0236794925, NA, -0.0…
#> $ trend_deaths_coeff_upr95_14d <dbl> 0.080093360, NA, 0.022194689, NA, 0.14063…
#> $ trend_deaths_preds_14d       <list> [<tbl_df[12 x 6]>], <NULL>, [<tbl_df[12 …
#> $ dt_deaths_est_14d            <dbl> 14.988415, NA, NA, NA, 9.903671, NA, NA, …
#> $ dt_deaths_lwr_14d            <dbl> 8.654240, NA, NA, NA, 4.928621, NA, NA, N…
#> $ dt_deaths_upr_14d            <dbl> 55.90936, NA, NA, NA, -1051.30328, NA, NA…
#> $ date_start_30d               <date> 2021-04-25, 2021-04-25, 2021-04-25, 2021…
#> $ date_end_30d                 <date> 2021-05-24, 2021-05-24, 2021-05-24, 2021…
#> $ cases_30d                    <dbl> 7545, 1806, 6371, 545, 7131, 32, 716263, …
#> $ trend_cases_30d              <chr> "Increasing", "Decreasing", "Likely decre…
#> $ trend_cases_coeff_30d        <dbl> 0.020014257, -0.060730704, -0.005639469, …
#> $ trend_cases_coeff_lwr95_30d  <dbl> 0.0027712747, -0.0693098542, -0.013201830…
#> $ trend_cases_coeff_upr95_30d  <dbl> 0.037257239, -0.052151554, 0.001922893, -…
#> $ trend_cases_preds_30d        <list> [<tbl_df[28 x 6]>], [<tbl_df[28 x 6]>], …
#> $ dt_cases_est_30d             <dbl> 34.63267, NA, NA, NA, 46.68539, NA, 36.74…
#> $ dt_cases_lwr_30d             <dbl> 18.60436, NA, NA, NA, 29.36814, NA, 24.53…
#> $ dt_cases_upr_30d             <dbl> 250.11854, NA, NA, NA, 113.77279, NA, 73.…
#> $ deaths_30d                   <dbl> 240, 73, 228, 3, 154, 11, 13006, 393, 0, …
#> $ cases_inc_30d                <dbl> 19.411784, 65.735786, 14.569023, 849.7170…
#> $ deaths_inc_30d               <dbl> 0.61758833, 2.65877145, 0.52145722, 4.717…
#> $ trend_deaths_30d             <chr> "Stable", "Decreasing", "Decreasing", NA,…
#> $ trend_deaths_coeff_30d       <dbl> 0.0044184199, -0.0243032375, -0.019043692…
#> $ trend_deaths_coeff_lwr95_30d <dbl> -0.0057519741, -0.0358803712, -0.02415841…
#> $ trend_deaths_coeff_upr95_30d <dbl> 0.014588814, -0.012726104, -0.013928973, …
#> $ trend_deaths_preds_30d       <list> [<tbl_df[28 x 6]>], [<tbl_df[28 x 6]>], …
#> $ dt_deaths_est_30d            <dbl> NA, NA, NA, NA, 18.40781, NA, 74.13533, N…
#> $ dt_deaths_lwr_30d            <dbl> NA, NA, NA, NA, 12.74496, NA, 37.34709, N…
#> $ dt_deaths_upr_30d            <dbl> NA, NA, NA, NA, 33.12665, NA, 4954.40056,…

# country with fastest growing case numbers in last 14 days
(iso <- df_trends %>% 
  filter(trend_cases_coeff_14d == max(trend_cases_coeff_14d, na.rm = TRUE)) %>% 
  pull(iso_a3))
#> [1] "TWN"
df_country <- df_covid %>% filter(iso_a3 == iso)
df_trends_country <- df_trends %>% filter(iso_a3 == iso)

country_plot(df_country, df_trends_country, add_title = TRUE)

epicentre-msf/covidutils documentation built on July 22, 2021, 1:03 p.m.