Using ILSAfile.info()
we can identify different
populations within an ILSA.
Moreover, we can also use this information for loading an specific population into the memory. Using the ‘REDS’ 2021 example data provided by this package:
We can decide to load a population, like BCGV1
, using
justload()
:
loaded <- justload(inputdir = system.file("extdata/reds", package = "ILSAmerge"),
population = "BCGV1")
This will result in a list of elements loaded by haven
,
in our example, a list of 11 elements:
class(loaded)
## [1] "list"
length(loaded)
## [1] 4
loaded[[1]]
## # A tibble: 10 × 314
## IDCNTRY IDSCHOOL ITLANGC IP1G00A IP1G00B IP1G00C
## <dbl+lbl> <dbl+lb> <dbl+lbl> <dbl+lbl> <dbl+lbl> <dbl+lbl>
## 1 784 [United Arab E… 1001 53 [Arabic] 2 [Mid] 3 [March] 9 (NA) [Omi…
## 2 784 [United Arab E… 1002 53 [Arabic] 1 [Early] 2 [February] 3 [Late]
## 3 784 [United Arab E… 1003 53 [Arabic] 1 [Early] 2 [February] 3 [Late]
## 4 784 [United Arab E… 1004 1 [English] 3 [Late] 2 [February] 3 [Late]
## 5 784 [United Arab E… 1005 53 [Arabic] 1 [Early] 3 [March] 3 [Late]
## 6 784 [United Arab E… 1006 53 [Arabic] 2 [Mid] 2 [February] 2 [Mid]
## 7 784 [United Arab E… 1007 53 [Arabic] 1 [Early] 3 [March] 3 [Late]
## 8 784 [United Arab E… 1008 53 [Arabic] 1 [Early] 3 [March] 2 [Mid]
## 9 784 [United Arab E… 1009 53 [Arabic] 2 [Mid] 3 [March] 1 [Early]
## 10 784 [United Arab E… 1010 53 [Arabic] 2 [Mid] 6 [June] 3 [Late]
## # ℹ 308 more variables: IP1G00D <dbl+lbl>, IP1GIAA <dbl+lbl>,
## # IP1GIAB <dbl+lbl>, IP1GIAC <dbl+lbl>, IP1GIAD <dbl+lbl>, IP1GIAE <dbl+lbl>,
## # IP1GIAF <dbl+lbl>, IP1GIAG <dbl+lbl>, IP1GIBA <dbl+lbl>, IP1GIBB <dbl+lbl>,
## # IP1GIBC <dbl+lbl>, IP1GIBD <dbl+lbl>, IP1G01A <dbl+lbl>, IP1G01B <dbl+lbl>,
## # IP1G01C1 <dbl+lbl>, IP1G01C2 <dbl+lbl>, IP2G01A1 <dbl+lbl>,
## # IP1G01AA <dbl+lbl>, IP1G02A <dbl+lbl>, IP1G02B <dbl+lbl>,
## # IP1G02C <dbl+lbl>, IP1G02D <dbl+lbl>, IP1G02E <dbl+lbl>, …
For some purposes, we might need to load only the column information,
most probably to check if the attributes are correct. We can do this by
setting justattributes = TRUE
:
loaded <- justload(inputdir = system.file("extdata/reds", package = "ILSAmerge"),
population = "BCGV1", justattributes = TRUE)
This will load all 11 datasets with 0 rows each:
length(loaded)
## [1] 4
loaded[[1]]
## # A tibble: 0 × 314
## # ℹ 314 variables: IDCNTRY <dbl+lbl>, IDSCHOOL <dbl+lbl>, ITLANGC <dbl+lbl>,
## # IP1G00A <dbl+lbl>, IP1G00B <dbl+lbl>, IP1G00C <dbl+lbl>, IP1G00D <dbl+lbl>,
## # IP1GIAA <dbl+lbl>, IP1GIAB <dbl+lbl>, IP1GIAC <dbl+lbl>, IP1GIAD <dbl+lbl>,
## # IP1GIAE <dbl+lbl>, IP1GIAF <dbl+lbl>, IP1GIAG <dbl+lbl>, IP1GIBA <dbl+lbl>,
## # IP1GIBB <dbl+lbl>, IP1GIBC <dbl+lbl>, IP1GIBD <dbl+lbl>, IP1G01A <dbl+lbl>,
## # IP1G01B <dbl+lbl>, IP1G01C1 <dbl+lbl>, IP1G01C2 <dbl+lbl>,
## # IP2G01A1 <dbl+lbl>, IP1G01AA <dbl+lbl>, IP1G02A <dbl+lbl>, …