A function to capture a set of import specifications for a
directory of data files. These specs can be used on the
libname function to correctly assign the data types for
imported data files. The
import engines will guess at the data types for any columns that
are not explicitly defined in the import specifications. Import
specifications are defined with the import_spec function.
The import spec syntax is the same for all data engines.
Note that the na and trim_ws parameters on the specs
function will be applied globally to all files in the library.
These global settings can be overridden on the import_spec
for any particular data file.
Also note that the specs collection is defined as an object
so it can be stored and reused.
See the write.specs and read.specs functions
for additional information on saving specs.
specs(..., na = c("", "NA"), trim_ws = TRUE)Named input specs. The name should correspond to the file name,
without the file extension.
The spec is defined as an import_spec object. See the
import_spec function for additional information on
parameters for that object.
A vector of values to be treated as NA. For example, the
vector c('', ' ') will cause empty strings and single blanks to
be converted to NA values. For most file types,
empty strings and the string 'NA' ('', 'NA') are considered NA.
For SAS® datasets and transport files, a single blank and a single dot
c(" ", ".") are considered NA. The value of the
na parameter on the specs function can be overridden
by the na parameter on the import_spec function.
Whether or not to trim white space from the input data values.
Valid values are TRUE, and FALSE. Default is TRUE. The value of the
trim_ws parameter on the specs function can be overridden
by the trim_ws parameter on the import_spec function.
The import specifications object.
libname to create a data library,
dictionary for generating a data dictionary, and
import_spec for additional information on defining an
import spec.
Other specs:
import_spec(),
print.specs(),
read.specs(),
write.specs()
library(readr)
# Create temp path
tmp <- file.path(tempdir(), "mtcars.csv")
# Create data for illustration purposes
df <- data.frame(vehicle = rownames(mtcars), mtcars[c("mpg", "cyl", "disp")],
stringsAsFactors = FALSE)
# Kill rownames
rownames(df) <- NULL
# Add some columns
df <- datastep(df[1:10, ], {
recdt <- "10JUN1974"
if (mpg >= 20)
mpgcat <- "High"
else
mpgcat <- "Low"
if (cyl == 8)
cyl8 <- TRUE
})
df
# vehicle mpg cyl disp recdt mpgcat cyl8
# 1 Mazda RX4 21.0 6 160.0 10JUN1974 High NA
# 2 Mazda RX4 Wag 21.0 6 160.0 10JUN1974 High NA
# 3 Datsun 710 22.8 4 108.0 10JUN1974 High NA
# 4 Hornet 4 Drive 21.4 6 258.0 10JUN1974 High NA
# 5 Hornet Sportabout 18.7 8 360.0 10JUN1974 Low TRUE
# 6 Valiant 18.1 6 225.0 10JUN1974 Low NA
# 7 Duster 360 14.3 8 360.0 10JUN1974 Low TRUE
# 8 Merc 240D 24.4 4 146.7 10JUN1974 High NA
# 9 Merc 230 22.8 4 140.8 10JUN1974 High NA
# 10 Merc 280 19.2 6 167.6 10JUN1974 Low NA
# Save to temp directory for this example
write_csv(df, tmp)
## Start Example ##
# Define import spec
spcs <- specs(mtcars = import_spec(vehicle = "character",
cyl = "integer",
recdt = "date=%d%b%Y",
mpgcat = "guess",
cyl8 = "logical"))
# Create library
libname(dat, tempdir(), "csv", import_specs = spcs)
# $mtcars
# library 'dat': 1 items
# - attributes: csv not loaded
# - path: C:\Users\User\AppData\Local\Temp\RtmpqAMV6L
# - items:
# Name Extension Rows Cols Size LastModified
# 1 mtcars csv 10 7 9.3 Kb 2020-11-29 09:47:52
# View data types
dictionary(dat)
# # A tibble: 7 x 10
# Name Column Class Label Description Format Width Justify Rows NAs
# <chr> <chr> <chr> <chr> <chr> <lgl> <int> <chr> <int> <int>
# 1 mtcars vehicle character NA NA NA 17 NA 10 0
# 2 mtcars mpg numeric NA NA NA NA NA 10 0
# 3 mtcars cyl integer NA NA NA NA NA 10 0
# 4 mtcars disp numeric NA NA NA NA NA 10 0
# 5 mtcars mpgcat character NA NA NA 4 NA 10 0
# 6 mtcars recdt Date NA NA NA NA NA 10 0
# 7 mtcars cyl8 logical NA NA NA NA NA 10 8
# Clean up
lib_delete(dat)