|
|
# UNCTAD
|
|
|
|
|
|
## Provider
|
|
|
|
|
|
* **provider_name**: UNCTAD
|
|
|
* **provider_longname**: United Nation Conference on Trade and Development
|
|
|
* **provider URL**: https://unctad.org/en/Pages/Home.aspx
|
|
|
* **region**: World
|
|
|
* **terms of use**: we need to ask permission: http://unctadstat.unctad.org/EN/Copyright.html and https://shop.un.org/rights-permissions
|
|
|
* **approximate number of datasets**: 150
|
|
|
|
|
|
|
|
|
## Data accessibility
|
|
|
|
|
|
* **SDMX: (If yes, links and SDMX releases implemented**: No
|
|
|
* **REST API**: No
|
|
|
* **Bulk Download: (if yes, links and formats**: No
|
|
|
* **Account**: No
|
|
|
|
|
|
## Desired datasets
|
|
|
|
|
|
* **Description**: all datasets. One must emulate the JS to download
|
|
|
all data.
|
|
|
- download.py has been developed with Selenium. Files are save in XLS
|
|
|
- one dataset can contain several XLS file one by dimensions combination
|
|
|
|
|
|
## Data tree
|
|
|
|
|
|
* **Existence of a hierachy of datasets on web site**:
|
|
|
* **How to recover the information**: download.py provides the
|
|
|
hierarchy while traversing the web site
|
|
|
|
|
|
## Datasets
|
|
|
|
|
|
* **datasetCode**: initals of dataset names
|
|
|
* **datasetName**: Cell A1 (remove reference to period at end of
|
|
|
string, except if dataset is discountinued)
|
|
|
* **how to get release date**: download.py recovers it from web site
|
|
|
* **dataset docHref**: yes, linked should be provided by download.py
|
|
|
* **dataset notes**: no
|
|
|
* **dimension_list**: dimensions are on cell A4, C4, E4, ... and A7
|
|
|
and sometimes B7. Values must be recovered while parsing XLS file
|
|
|
* **use of attributes**: yes
|
|
|
* **attribute_list**: (provided or to be made up from the series)
|
|
|
* **available frequencies**: M, Q, A
|
|
|
* **availability of previous updates**: no
|
|
|
* **existence of real time datasets**: no
|
|
|
|
|
|
## Series
|
|
|
|
|
|
* **Series key**: initials from series name
|
|
|
* **Series name**: concatenate dimensions (dimensions on Row 4,
|
|
|
possible extra dimension in column B, regions (colume A)
|
|
|
* **Series docHref**: no
|
|
|
* **Series notes**: no
|
|
|
* **missing values**: '_' o '..'
|
|
|
* **date format**: yyyy, Qq yyyy, mmm. yyyy
|
|
|
* **mixed frequencies in the same dataset**: no
|
|
|
|
|
|
## Updates
|
|
|
|
|
|
* **calendar of future updates**: no
|
|
|
* **summary of previous updates**: no
|
|
|
* **regular updates**: no
|
|
|
* **RSS flow**:
|
|
|
* **best way to monitor updates**: check for update on web site
|
|
|
|
|
|
## Special problems
|
|
|
|
|
|
- need Selenium to go through the web site and download the files
|
|
|
- Chrome-headless seesm to work better than Firefox-headless
|
|
|
|
|
|
## Other remarks
|
|
|
|
|
|
|
|
|
|
|
|
|