bouldercodehub / rhdb Goto Github PK

jsonlite only suggests it, but we use function from jsonlite that relies on it.

mrid is returned as list for empty data frame case

Need two forms of error handling (at least):

1. When the http request returns the weird html. Should have tryCatch wrap it and try ~5 times before erroring out
2. When there is no data for a given sdi/timestep/mrid, the code fails, rather than returning NULL or similar

The following never gets data, and always returns an error request from API, but doesn't tell R user why

hdb_query(21327, "lc", "m", "2020-09", "2020-08", 3125)

from https://www.usbr.gov/uc/water/hydrodata/reservoir_data/meta.csv

add way to search through this, or get it as a df.

as should value

need more than just uc reservoir meta data. This does include some Mead, Mohave, and Havasu data, but only 4 variables.

Need something similar for LC (and others)

queries that result in no data error out instead of returning no data

ex:
hdb_query(21327, "lc", "m", "2007-08", "2009-08", 1684)

error is:
Error in tsdata[, c("sdi", "t", "mrid", "v", "units")] :
subscript out of bounds