Clean microdata to be used in PIP methods.
Details
md_clean_data() returns a list of elements whose main object is a
data.table with the necessary transformations to be included in PIP methods.
Data is available in element $data. The other elements provide the number
of observations that were modified depending on test performed. The name of
elements are in the form p_s, where p (or prefix) refers to the test and s
(the suffix) refers to the name of the variable evaluated.
Prefixes are:
nna: Number of NA in variable
nng: Number of negative values
ina: Index of obs with NA in variable
ing: Index of obs with negative values
Examples
# Load example data
data("md_GHI_2000_income")
# Clean microdata
res <- wbpip:::md_clean_data(
md_GHI_2000_income,
welfare = "welfare",
weight = "weight")
#> ℹ 2 negative values in variable "welfare" were dropped
#> ℹ Data has been sorted by variable "welfare"
res$data
#> country_code survey_year weight welfare area gender
#> <char> <num> <num> <num> <char> <char>
#> 1: GHI 2000 480.3053 0.00 rural female
#> 2: GHI 2000 1269.3970 0.00 urban male
#> 3: GHI 2000 1264.1021 0.00 rural male
#> 4: GHI 2000 928.6781 0.00 urban male
#> 5: GHI 2000 784.0803 0.00 urban female
#> ---
#> 1994: GHI 2000 786.4416 50647.29 urban female
#> 1995: GHI 2000 1882.1450 56400.00 urban male
#> 1996: GHI 2000 740.0855 60580.58 urban female
#> 1997: GHI 2000 1624.2939 132221.84 urban female
#> 1998: GHI 2000 1023.2650 188223.50 urban male