Right now, we have a few functions to "get" information. @CommonClimate suggested that the get_all functions should be grouped together with a field parameter (eg., dataset_name, dataset_id, archiveType...). See issue #39. Right now, this function returns unattached list and in the case of archiveType, the unique entries (which may be needed for search). So first question, should we only return unique entries in that list or all?
@alexkjames proposed to have a function that allows to build a queryable dataframe. Right now, it's limited to lat/lon but I could see being expanded pretty soon (see issue #40). In theory this would look like get_timeseries with more control over what is being expanded. One option could be create a get function with a field option that would append to a dataframe if used in method cascading.
If we decide not to return unique options for get_all should we collapse get_all and get into one call and build a dataframe progressively?
Right now, we have a few functions to "get" information. @CommonClimate suggested that the
get_all
functions should be grouped together with afield
parameter (eg., dataset_name, dataset_id, archiveType...). See issue #39. Right now, this function returns unattached list and in the case of archiveType, the unique entries (which may be needed for search). So first question, should we only return unique entries in that list or all?@alexkjames proposed to have a function that allows to build a queryable dataframe. Right now, it's limited to lat/lon but I could see being expanded pretty soon (see issue #40). In theory this would look like
get_timeseries
with more control over what is being expanded. One option could be create aget
function with afield
option that would append to a dataframe if used in method cascading.If we decide not to return unique options for
get_all
should we collapseget_all
andget
into one call and build a dataframe progressively?