chartplotter

The goal of chartplotter is to

  • plot the child’s growth curves on pre-defined charts;
  • plot two (visit) lines indicating a present and a future visit;
  • find matches from donor data similar to the target child;
  • predict the most likely future visit value;
  • represent the inherent uncertainty of the prediction.

Installation

You need access to growthcharts/donorloader in order to install chartplotter.

The following statements will install the chartplotter package, as well as any of its dependencies:

install.packages("remotes")
remotes::install_github("growthcharts/chartplotter")
remotes::install_github("growthcharts/jamesdemodata")

The main function in chartplotter is process_chart().

Example 1: Plot child’s data onto a growth chart 

library(chartplotter)
library(bdsreader)
library(svglite)

fn <- system.file("extdata", "bds_v2.0", "smocc", "Laura_S.json", package = "jamesdemodata")
tgt <- bdsreader::read_bds(fn)

# No need to manually set file path
svglite::svglite(file = "man/figures/chart1.svg", height = 29.7/2.54, width = 21/2.54)
g <- process_chart(tgt, chartcode = "NMBA")
grid::grid.draw(g)
dev.off()
#> quartz_off_screen 
#>                 2
Dutch girls, 0-4 years

Dutch girls, 0-4 years

Example 2: Predict height at 3y9m when child is 3 months

Suppose the baby is 3 months old, and that we want to predict the future child’s height at the age of 3y9m. The following examples finds 25 matches to the child, and plots the observed curves of those matches as grey curves.

The blue line indicates the predicted height at age 3y9m. The variation between the grey curves at age 3y9m indicates the amount of uncertainty of the prediction.

set.seed(61771)
svglite(file = "man/figures/chart2.svg", height = 29.7/2.54, width = 21/2.54)
g <- process_chart(tgt, chartcode = "NMBA", dnr = "2-4", period = c(0.25, 3.75), 
                   nmatch = 25, show_future = TRUE)
grid::grid.draw(g)
dev.off()
#> quartz_off_screen 
#>                 2
Dutch girls, 0-4 years

Dutch girls, 0-4 years

Example 3: Predict height at 3y9m when child is 2 years

Same as before, but now using all data up to (but not beyond) the age of 2 years. The variation between the grey curves at age 3y9m is now much smaller.

svglite(file = "man/figures/chart3.svg", height = 29.7/2.54, width = 21/2.54)
g <- process_chart(tgt, chartcode = "NMBA", dnr = "2-4", period = c(2.0, 3.75), 
                   nmatch = 25, show_future = TRUE)
grid::grid.draw(g)
dev.off()
#> quartz_off_screen 
#>                 2
Dutch girls, 0-4 years

Dutch girls, 0-4 years

Example 4: Square plot of height, chart NMBH 

svglite(file = "man/figures/chart4.svg", height = 18/2.54, width = 18/2.54)
g <- process_chart(tgt, chartcode = "NMBH", quiet = FALSE, dnr = "2-4",
                   period = c(2.0, 3.75), nmatch = 25,
                   show_future = TRUE, show_realized = TRUE)
#> chartcode:  NMBH
grid::grid.draw(g)
dev.off()
#> quartz_off_screen 
#>                 2
Dutch girls, 0-4 years

Dutch girls, 0-4 years