Mészáros M (2023). restatapi: Search and Retrieve Data from Eurostat Database. -R package version 0.20.1, https://github.com/eurostat/restatapi. +R package version 0.20.3, https://github.com/eurostat/restatapi.
@Manual{,
title = {restatapi: Search and Retrieve Data from Eurostat Database},
author = {Mátyás Mészáros},
year = {2023},
- note = {R package version 0.20.1},
+ note = {R package version 0.20.3},
url = {https://github.com/eurostat/restatapi},
}
diff --git a/docs/index.html b/docs/index.html
index 8cfa027..e4aa90e 100644
--- a/docs/index.html
+++ b/docs/index.html
@@ -37,7 +37,7 @@
@@ -72,9 +72,12 @@
Version 0.20.0 enables all the functionality for the new dissemination chain. It has breaking changes concerning the date_filter as in the old dissemination the value was assigned to the first day of the month, quarter and year so it was enough to filter for one day to get the value. Under the new API the value belongs to the full period. If a date range does not cover the whole period no values will be returned. E.g. to get the value of the whole quarter the date filter should start at least on the first date of the quarter and end at least on the last day of the quarter. With concrete numbers to get the value for 2022/Q3, the startDate should be 2022-07-01 or earlier and the endDate 2022-09-30 or later. In the old version it was enough if the period included the day 2022-07-01 only.
Version 0.20.0 enables all the functionality for the new dissemination chain and from version 0.20.3 it is the default API.
+The new API has breaking changes concerning the date_filter, as in the old dissemination the value was assigned to the first day of the month, quarter and year so it was enough to filter for one day to get the value. Under the new API the value belongs to the full period. If a date range does not cover the whole period no values will be returned. E.g. to get the value of the whole quarter the date filter should start at least on the first date of the quarter and end at least on the last day of the quarter. With concrete numbers to get the value for 2022/Q3, the startDate should be 2022-07-01 or earlier and the endDate 2022-09-30 or later. In the old version it was enough if the period included the day 2022-07-01 only.
In addition, if the date filter is only one day (e.g. startDate=2007-07-02&endDate=2007-07-02) then the new API will give you back the values for all the time periods in the dataset applying the filter for the other concepts. But if the time period changes to more than one day (e.g. startDate=2007-07-01&endDate=2007-07-02) then the new API gives back only those values which are covered by the range.
options(restatapi_update=TRUE)
options(restatapi_cache_dir=file.path(tempdir(),"restatapi"))Example 6: First download the annual (select_freq="A") air passenger transport data for the main airports of Montenegro (avia_par_me) and do not cache any of the data (cache=FALSE). Then from the same table download the monthly (select_freq="M") and quarterly (filters="Q...) data for 2 specific airport pairs/routes (filters=...ME_LYPG_HU_LHBP+ME_LYTV_UA_UKKK") in August 2016 and on 1 July 2017 (date_filter=c("2016-08","2017-07-01")). The filters are provided in the format how it is required by the REST SDMX web service. Then download again the monthly and quarterly data (filters=c("Quarterly","Monthly")) where there is exact match in the DSD for “HU” for August 2016 and 1 March 2014 (date_filter=c("2016-08","2014-03-01")). This query will provide only monthly data for 2016, as the quarterly data is always assigned to the first month of the quarter and there is no data for 2014. Since there is no exact match for the “HU” pattern, it will return all the monthly data for August 2016 and put the labels (like the name of the airports and units) so the data can be easier understood (label=TRUE). Finally, download only the quarterly data (select_freq="Q") for several time periods (date_filter=c("2017-03",2016,"2017-07-01",2012:2014), the order of the dates does not matter) where the “HU” pattern can be found anywhere, but only in the code column of the DSD (filters="HU",exact_match=FALSE,name=FALSE). The result will be all the statistics about flights from Montenegro to Hungary in the 3rd quarter of 2017, as there is no information for the other time periods.
Example 6: First download the annual (select_freq="A") air passenger transport data for the main airports of Montenegro (avia_par_me) and do not cache any of the data (cache=FALSE). Then from the same table download the monthly (select_freq="M") and quarterly (filters="Q...) data for 2 specific airport pairs/routes (filters=...ME_LYPG_HU_LHBP+ME_LYTV_UA_UKKK") in August 2016 and on 1 July 2017 (date_filter=c("2016-08","2017-07-01")). The filters are provided in the format how it is required by the REST SDMX web service. Under the old API it returned the value for the selected routes for the month August 2016, July 2017 and the 3rd quarter of 2017. Meanwhile under the new API it returns all the quarterly and monthly value as there is a single day in the date_filter. Then download again the monthly and quarterly data (filters=c("Quarterly","Monthly")) where there is exact match in the DSD for “HU” for August 2016 and 1 March 2014 (date_filter=c("2016-08","2014-03-01")). This query will provide only monthly data for 2016, as the quarterly data is always assigned to the first month of the quarter and there is no data for 2014. Since there is no exact match for the “HU” pattern, it returned all the monthly data for August 2016 and put the labels (like the name of the airports and units) so the data can be easier understood (label=TRUE) under the old API. Under the new API it returns all the quarterly and monthly data as there is a single day in the date_filter. Finally, download only the quarterly data (select_freq="Q") for several time periods (date_filter=c("2017-03",2016,"2017-07-01",2012:2014), the order of the dates does not matter) where the “HU” pattern can be found anywhere, but only in the code column of the DSD (filters="HU",exact_match=FALSE,name=FALSE). The result was all the statistics about flights from Montenegro to Hungary in the 3rd quarter of 2017, as there is no information for the other time periods under the old API. Under the new API it will give back all the quarterly data in dataset for flights from Montenegro to Hungary because in the date_filter there is a single day.
dt<-get_eurostat_data("avia_par_me",select_freq="A",cache=FALSE)
dt<-get_eurostat_data("avia_par_me",
diff --git a/docs/news/index.html b/docs/news/index.html
index bb5bf45..fbd812b 100644
--- a/docs/news/index.html
+++ b/docs/news/index.html
@@ -17,7 +17,7 @@
NEWS.md
option(restatapi_verbose=TRUE) used for the extract_data(), extract_dsd(), get_compressed_sdmx(),get_eurostat_dsd(), get_eurostat_raw(), get_eurostat_bulk(), get_eurostat_data()and get_eurostat_toc() functionsget_eurostat_dsd() and extract_dsd() function for the new API in version 0.14.0 to variable in the .restatapi_env environmentget_compressed_sdmx() function when format="gz" is usedget_eurostat_bulk() function was adjusted for the changing letter case under the new APIget_eurostat_raw() function when the check_toc=TRUE option used under the new APIextract_data() function for the new API under Windowscreate_filter_table() function was adjusted for the changing letter casecreate_filter_table() function was adjusted for the changing letter case under the new APIextract_data() function to handle files which was downloaded using the link from the TOC (as those links are still referring to the old API)extract_toc() function returns a data.table and not a character vector to have correct number of columns for all casesget_eurostat_toc() function was changed to handle the different format from the extract_toc() functionget_eurostat_data() function was adjusted to handle the changing DSD contentget_eurostat_dsd() function was changed to filter out only the relevant concepts and code from the new general DSD filesget_eurostat_raw() function was changed to handle the different compression formats in the get_compressed_sdmx() functionget_eurostat_raw(), get_eurostat_bulk() and get_eurostat_date() functions were adjusted to handle the changing letter case and dimension namesget_eurostat_raw(), get_eurostat_bulk() and get_eurostat_data() functions were adjusted to handle the changing letter case and dimension namesget_eurostat_data() function when select_freq was used with only 1 filtertime_filter and availability of data under the new APIextract_dsd(concept = NULL, dsd_xml = NULL, lang = "en", api_version)extract_dsd(concept = NULL, dsd_xml = NULL, lang = "en")a character string either en, de or fr to define the language version for the name column
of the DSD. It is used only in the new API. The default is en - English.
the value of the API_VERSIONING from the config file