The following is questions posed by a user and my answers to them - putting here so the conversation is discoverable
Q: When you submit a query to download data (´occ_download()´), is the absolute limit 200.000 observations, disregarding how high the taxon key you are using (i.e. if many species included)?
A: No, there’s no upper limit for the download API.
Q: When parallelizing the query for this example
chunksof10 ← split(keys, ceiling(seq_along(keys)/10))
data ← llply(chunksof10, occ_search,… limit=10.000)
> Is the limit = 10.000 (e.g.) the max number of observations obtained for each species in each "chunk" of ten species, or for the complete chunk of species?
A: `occ_search()` internally does paging, so 10,000 is not a GBIF imposed limit, but just a number I picked for purposes of the example. The 10,000 number would be for each run/iteration of occ_search within the lapply call
> Q: Is there any way you know you can directly get count data (for observations and number of species) over a grid or raster as the ones shown for complete datasets http://www.gbif.org/species/5231103 ?
A:
* first, that used to be possible in the old GBIF API, but was removed. However, they've recently added it back into their API, and we will soon incorporate https://github.com/ropensci/rgbif/issues/215
* second, you can use the GBIF maps service at http://www.gbif.org/developer/maps - but that only gives you maps good for web applications - or static images