farrago

{farrago} is an R package serving as a collection of tools for data workflows and analysis, with focus on health surveillance data. Although {farrago} primarily serves as a personal collection of odds-and-ends picked-up or created over the past several years, it may assist wider audiences as well. The package is organized by general purpose/functionality, which may eventually be separated into discrete packages.

Installation

{farrago} is only available from GitHub and the latest version can be installed with:

# install.packages("devtools")
devtools::install_github("al-obrien/farrago")

Use-cases

{farrago} has a variety of functions available. They are roughly organized into the following categories:

1. Calculations: provides algorithms for some routine processes such as…

   • Determining pregnancy trimesters

   • Assigning episode periods (e.g. for repeat infections)

   • Collapsing time-steps

   • Determining overlaps in time

   • Basic metrics such as rates, max/min, etc.

2. Conversions: helper functions to convert between common formats in epidemiology

   • Replace all blank values to NA (e.g. when importing data from SAS)

   • Switch between flu and calendar weeks

   • Determine flu season from date

   • Quickly convert a table to image (png)

   • Basic conversions such as from numbers to percent, number to factor, etc.

3. Creation: generate new content

   • Make multi-level factors similar to SAS ‘multi-label’ functionality

   • Create hypercubes (i.e. n-dimensional table including group summaries and totals)

   • Determine break points from set of values

4. Transferal: methods to move objects and data

   • Easily stow() and retrieve() data-sets to make efficient use of RAM

   • File transfer using WinSCP wrapper

   • Locate files

   • Pass code and retrieve data from SAS (primarily for use with Classic 9.4)

5. Plotting: helper functions for shared legends and less common plots such as bulls-eye charts and X-splines

6. Miscellaneous

Example

This is a basic example using a sub-set of functions from {farrago}…

# Load libraries
library(farrago)
library(magrittr)
library(dplyr)
library(lubridate)

# Download from configured SFTP location
transfer_winscp(file ='my_rmt_file.csv'),
               direction = 'download',
               connection = 'sftp://myusername:[email protected]/'
               rmt_path = './location/',
               drop_location = 'C:/PATH/TO/DESIRED/FOLDER/')

# Non-sense data for example
my_rmt_file <- tibble::tribble(~grp_id, ~date, ~date_of_birth, ~condition, ~date_of_birth_child, 
                               1, '2020-01-01', '1970-06-04', 'alive', '1991-01-01',
                               1, '2020-01-01', '1980-04-05', '', '1990-02-04',
                               1, '2020-01-03', '1930-04-05', 'alive', '',
                               1, '2020-01-04', '1967-04-05', 'alive', '1998-01-21',
                               2, '2020-01-01', '1978-04-05', 'alive', '1998-06-21',
                               2, '2020-09-10', '1970-04-05', 'alive', '1992-09-13',
                               2, '2020-09-21', '1949-04-05', 'dead', '1987-01-03',
                               3, '2020-01-01', '1977-04-05', '', '1992-01-21',
                               3, '2020-01-02', '1944-04-05', 'alive', '',
                               3, '2020-01-21', '1943-06-05', 'alive', '1967-09-12',
                               3, '2020-01-22', '1969-07-05', 'alive', '2006-12-21',
                               3, '2020-04-22', '', NA, NA,
                               3, '2021-06-09', '1978-09-21', 'dead', '1992-01-21') %>%
  dplyr::mutate_at(vars(contains('date')), ymd)

# Remove blanks
my_rmt_file <- convert_blank2NA(my_rmt_file)

# Determine episode period based on first date by group
my_rmt_file$episode <- assign_episode(data = my_rmt_file,
                                      grp_id = grp_id,
                                      date = date,
                                      threshold = 10)

# Determine age and age group from date
my_rmt_file$age <- calculate_age(my_rmt_file$date_of_birth)
my_rmt_file$age_grp <- create_breaks(my_rmt_file$age, breaks = c(0, 10, 20, 30, 40, 50, 60, 70, 80, 90), format = TRUE)
 
# Calculate trimester based on dob (of child)
my_rmt_file <- calculate_trimesters(my_rmt_file, date_of_birth_child)
#> Warning in calculate_trimesters(my_rmt_file, date_of_birth_child): No variable
#> for gestation length was provided, all pregnancies will assume the average
#> pregnancy length of: 40

# View final dataset
knitr::kable(my_rmt_file)

grp_id	date	date_of_birth	condition	date_of_birth_child	episode	age	age_grp	tri1_s	tri1_e	tri2_s	tri2_e	tri3_s	preterm
1	2020-01-01	1970-06-04	alive	1991-01-01	1	51	50-59	1990-03-27	1990-06-26	1990-06-27	1990-09-26	1990-09-27	0
1	2020-01-01	1980-04-05	NA	1990-02-04	1	41	40-49	1989-04-30	1989-07-30	1989-07-31	1989-10-30	1989-10-31	0
1	2020-01-03	1930-04-05	alive	NA	1	91	>=90	NA	NA	NA	NA	NA	NA
1	2020-01-04	1967-04-05	alive	1998-01-21	1	54	50-59	1997-04-16	1997-07-16	1997-07-17	1997-10-16	1997-10-17	0
2	2020-01-01	1978-04-05	alive	1998-06-21	1	43	40-49	1997-09-14	1997-12-14	1997-12-15	1998-03-16	1998-03-17	0
2	2020-09-10	1970-04-05	alive	1992-09-13	2	51	50-59	1991-12-08	1992-03-08	1992-03-09	1992-06-08	1992-06-09	0
2	2020-09-21	1949-04-05	dead	1987-01-03	3	72	70-79	1986-03-29	1986-06-28	1986-06-29	1986-09-28	1986-09-29	0
3	2020-01-01	1977-04-05	NA	1992-01-21	1	44	40-49	1991-04-16	1991-07-16	1991-07-17	1991-10-16	1991-10-17	0
3	2020-01-02	1944-04-05	alive	NA	1	77	70-79	NA	NA	NA	NA	NA	NA
3	2020-01-21	1943-06-05	alive	1967-09-12	2	78	70-79	1966-12-06	1967-03-07	1967-03-08	1967-06-07	1967-06-08	0
3	2020-01-22	1969-07-05	alive	2006-12-21	2	52	50-59	2006-03-16	2006-06-15	2006-06-16	2006-09-15	2006-09-16	0
3	2020-04-22	NA	NA	NA	3	NA	NA	NA	NA	NA	NA	NA	NA
3	2021-06-09	1978-09-21	dead	1992-01-21	4	43	40-49	1991-04-16	1991-07-16	1991-07-17	1991-10-16	1991-10-17	0

# Save file for easy retrieval later
my_rmt_file_stowed <- stow(my_rmt_file, cleanup = TRUE)