Load a bundled dataset by slug

Reads the table's CSV and applies the column names and classes from the bundled schema, so the result is the same typed data frame on every platform regardless of how read.csv() would have guessed. Every table also ships as a Parquet file holding the same typed frame; format = "parquet" reads that copy instead. The first load of a table is cached for the session; later calls return the cached copy unless refresh = TRUE.

Usage

mdload(slug, refresh = FALSE, format = c("csv", "parquet"))

Arguments

slug

Dataset slug; see the slug column of mdcat().

refresh

Re-read the file, and check it against the signed manifest again, even if a cached copy exists.

format

"csv" (default) or "parquet": which shipped copy to read. Both give the same columns, with the same names and types, and the same values, with one exception: in the double columns of five tables (tps_assault_sample and its copy samples_tps_assault_sample, vic_criminal_incidents, vic_lga_criminal_incidents and vic_recorded_offences) one or two values of the Parquet copy differ from the CSV's in the last binary digit (a relative difference below 3e-16). The CSV text read back gives the double nearest to it; those Parquet copies were written from values computed before the CSV text was, and hold the next double.

Value

A data frame.

Examples

d <- mdload("arsau_2023_uof_main_records")
str(d[, 1:4])
#> 'data.frame':	5 obs. of  4 variables:
#>  $ IncidentYear     : int  2024 2024 2024 2024 2024
#>  $ BatchFileName    : chr  "SYNTHETIC-FIXTURE-001" "SYNTHETIC-FIXTURE-002" "SYNTHETIC-FIXTURE-003" "SYNTHETIC-FIXTURE-004" ...
#>  $ PoliceServiceType: chr  "Municipal Police Service" "Municipal Police Service" "Provincial Police Service" "Municipal Police Service" ...
#>  $ PoliceService    : chr  "Sample Regional Police Service" "Sample Regional Police Service" "OPP" "Sample Regional Police Service" ...