I’m working on a project that uses the rbison and rgbif, we are mostly focused on plants data from samples collected in the US. As someone new to programming and R these packages have been extremely helpful.
I noticed some differences in the rbison and rgbif packages and was wondering if there were going to be any efforts to improve the rbison package? I ran into some trouble the other day when I attempted to use the bison_solr() function to page through BISON results and ended up sending repeated nonsense to the interface (my fault I should have been more careful) resulting in degraded performance of their Solr interface. The occ_data() and occ_search() functions in the rgbif package handle requests internally and has the added feature of a hard maximum of 200,000, which prevents new users like me from doing foolish things. Is this something that will be implemented in the rbison package?
I ran into some trouble the other day when I attempted to use the bison_solr() function to page through BISON results and ended up sending repeated nonsense to the interface (my fault I should have been more careful) resulting in degraded performance of their Solr interface.
Can you give more details. What do you mean by nonsense, and how did you know there was degraded performance.
The occ_data() and occ_search() functions in the rgbif package handle requests internally and has the added feature of a hard maximum of 200,000, which prevents new users like me from doing foolish things. Is this something that will be implemented in the rbison package?
rgbif is a pkg for GBIF and rbison for USGS"s BISON service. The infrastructure is different between GBIF and BISON. So we can’t necessarily do the same stuff with the two data sources. Yes, you’re right that occ_data/occ_search do internal paging - that is, GBIF has a max of 300 records per request, which is quite small, so we essentially do a for loop to get all the records you want up to 200K. After that, use the download API through functions that start with occ_download
I had forgotten to set the limit for scientific notation in R so when I asked for a page in Solr R would place 1e-05 instead of a number into the URL. We have been working closely with he USGS BISON staff and they informed me of the issue when it degraded the performance of their site.
