INEGI has published a breaking change with the current inegiR version (error: "d_dates does not exist"
) - fix has been provided by a user here. However, I still need to test and send an update to CRAN (Work in progress with no ETA yet).
If you are eager for new version, please use devtools to update the package:
library(devtools)
devtools::install_github(repo = "Eflores89/inegiR")
Version 3 of inegiR is now on CRAN. This is a big update that breaks with most functionality in previous versions.
Most importantly, this version is now (finally) compatible with INEGI's updated API (v2.0). Similar to previous versions, you can use the inegi_series function to download the data series you want. Now, you just need to add the indicador code.
library(inegiR)
token_inegi <- "your-own-token"
gdp <- inegi_series(serie = "381016", token = token_inegi)
tail(gdp)
date | date_shortcut | values | notes |
---|---|---|---|
2016-01-01 | Q1 | 14080629 | NA |
2016-04-01 | Q2 | 14333716 | NA |
2016-07-01 | Q3 | 14421873 | NA |
2016-10-01 | Q4 | 15012430 | NA |
2017-01-01 | Q1 | 14471824 | NA |
2017-04-01 | Q2 | 14586819 | NA |
There is an added option to download the data in tibbletime format.
gdp <- inegi_series(serie = "381016", token = token_inegi, as_tt = TRUE)
class(gdp)
[1] "tbl_time" "tbl_df" "tbl" "data.frame"
Like previous version, you can also download metadata. However, INEGI's new API does not explicitly serve the "human" metadata in the call. Instead, you get some codes and access to catalogs to describe those codes. For example, for the same data series...
gdp <- inegi_series(series_id = "381016", token = token_inegi, metadata = TRUE)
t(gdp$metadata)
source | 17 |
topic | 1160165000100010 |
notes | 2830 |
last_update | 22/08/2017 02:36:48 p. m. |
region | 00 |
units | 278 |
indicator_ID | 381016 |
frequency | 6 |
call_local_time | 2019-07-06 14:56:02 |
call_unmasked | https://www.inegi.org.mx/app/api/indicadores/desarrolladores/jsonxml/INDICATOR/381016/en/00/false/BIE/2.0/your-token-will-be-here?type=json |
Beware, not all the date formats are supported (only anual, trimestral, monthly and biweekly data) and there is a new date_shortcut column, which makes it easier to see the frequency of the series.
To obtain the code descriptions, there are some new functions, which begin with incat_:
# obtain the frequency codes.
frequencies <- incat_freq(token_inegi)
All of specific indicator functions (i.e. unemployment rate) have been rewritten and now start with inind_.
There is a new function for downloading a vector of more than one series in one go. They do not have to be the same frequency. For example, GDP and unemployment.
series_needed <- c("381016", "444612")
series_names <- c("GDP - old series", "Unemployment rate")
data_for_project <- inegi_series_multiple(series_id = series_needed,
token = token_inegi,
names = series_names)