psrcctpp tips & tricks • psrcctpp

Querying by variable

In addition to requesting entire tables, the package/API can return individual variables. If you prefer this, you’ll need to identify the variable codes, so naturally there’s a ctpp_varsearch function to assist; the only parameter is the table code in question. Once these codes are identified, their character vector can be used as the second argument (table_code) get_psrc_ctpp().

shhh <- suppressPackageStartupMessages
shhh(library(psrcctpp))
shhh(library(magrittr))
shhh(library(dplyr))
shhh(library(stringr))

ctpp_varsearch("A101101") %>% head()

##          name           label
##        <char>          <char>
## 1: A101101_e2  Under 16 years
## 2: A101101_e1 Total, all ages
## 3: A101101_e3 16 and 17 years
## 4: A101101_e4  18 to 20 years
## 5: A101101_e5  21 to 24 years
## 6: A101101_e6  25 to 34 years

x <- get_psrc_ctpp(scale="tract", 
                   table_code=c("A101101_e1","A101101_e2"), # Variable list instead of table
                   dyear=2016)

mutate(x, res_label=str_sub(res_label, 7L, 14L),           # abbreviate to fit in frame
          category =str_sub(category, 1L, 15L)) %>%
  select(c(3,6:8)) %>% head()

##    res_label        category estimate estimate_moe
##       <char>          <char>    <num>        <num>
## 1:   Tract 1 Total, all ages     7260          481
## 2:   Tract 1  Under 16 years     1120          264
## 3:   Tract 2 Total, all ages     7900          437
## 4:   Tract 2  Under 16 years     1150          196
## 5:   Tract 3 Total, all ages     2830          223
## 6:   Tract 3  Under 16 years      555          138

Bypassing the API

In the unlikely event of API downtime, an older method exists in the package to read CTPP ftp files pre-downloaded to a network location (it takes too long to download the entire ftp zipfile for each call). To use this fallback option, specify get_psrc_ctpp( ..., filepath="default") while on PSRC VPN. It is also possible to go completely offline by saving a full copy to a local directory, then specifying that directory in place of "default". There is no API-free fallback for the table and variable lookup functions, however.

y <- get_psrc_ctpp("tract", "A101101", 2016, filepath="default")

mutate(y, res_label=str_sub(res_label, 7L, 14L),    # abbreviate to fit in frame
          category =str_sub(category, 1L, 15L)) %>%
  select(c(3,6:8)) %>% head()

##    res_label estimate estimate_moe category
##       <char>    <num>        <num>   <char>
## 1:   Tract 1     7260          481    Total
## 2:   Tract 2     7900          437    Total
## 3:   Tract 3     2830          223    Total
## 4:   Tract 4     6260          500    Total
## 5:   Tract 4     5165          481    Total
## 6:   Tract 5     3155          182    Total