matthias-da
commented
2 years ago Thank you, Daniel. Sorry for the late reply. Its now on the way to the task view.
Closed matthias-da closed 2 years ago
matthias-da
commented
2 years ago Thank you, Daniel. Sorry for the late reply. Its now on the way to the task view.
From Daniel Antal daniel.antal@dataobservatory.eu | Daniel Antal daniel.antal@dataobservatory.eu | 1:34 AM (14 hours ago) | |
Dear Matthias,
I have read some of your articles, but your CRAN Task View did not come across before. I'd like to recommend a few packages that may be interesting for you.
The eurostat package that is in your review (from rOpenGov) gave birth to two further packages. Sub-national boundaries change in Europe often in several hundred places over 2-3 years, and unfortunately regional coding does not keep up with this pace in statistical metadata. Eurostat's regional data services usually cannot be organized into a panel, because the regional codes (and labels) do not match. The regions package provides a patch between 1999 and 2021.
Today statcodelists was released on CRAN, which aims to make the work of the regions package more general and facilitate the use of SDMX metadata codelists in R. It is a small step, but I would like to add new functionality to the Eurostat, OECD and other packages, and I am planning a reproducible workflow to publish dataset after analysis in R on Zenodo, the European open science repository that matches SDMX, W3C and DataCite metadata standards -- making the results in terms of data tables ( not notebooks, or written descriptions) readily publishable, and interoperable with a wide range of semantic applications, like the EU Open Data Portal, or cross-synchronization with statistical agencies.
The iotables makes the most structured statistical products, the input-output tables tidy, and allows for economic and environmental impact analysis with formatting the data received from the Eurostat data warehouse into appropriate, validated, matrix forms. I am planning to extend this soon to work with the OECD package, too, in your review. The package documentation currently reproduces the entire Eurostat manual on IOTs, and started to reproduce the UN Statistics one, too.
The retroharmonize package aims to give a reproducible workflow for harmonizing the variable names, variable labels and coding schemas of raw survey files. We are currently testing it on Eurostat microdata (via academic account) and waiting for official access to microdata to build more use cases, but the package works with Eurobarometer, Arab baromter and Afrobarometer.
Best regards,
Daniel Antal, CFA co-founder Digital Music Observatory +31615058460 (tel) +36305150534 (viber & whatsapp) Follow us on || LinkedIn || Twitter || Zotero (data) || Github (open code) ||