Closed brownag closed 11 months ago
Turns out it was a fairly simple fix. The same type of problem in both functions: creating dates in the current year rather than in the target year sometimes leads to offsets in julian days due to leap years. This error cropped up right on 2024-01-01.
This means that the values returned by summarizeSoilTemperature()
have been, up until this point, slightly dependent on the year in which they are calculated (regardless of the input data). Now the year is derived from the input data, so it should be stable across years. Tthe expected values for winter and summer temperature have been updated (change in the hundreths place) and tolerance added. The tests are all now skipped on CRAN.
Nice catch! Thanks for fixing this.
fetchHenry()
andwaterDayYear()
are failing tests due to slight offsets in the expected values. I will need to resubmit the package within two weeks, so unless there is an obvious fix I will likely skip these tests on CRAN going forward to prevent minor issues like this causing problems for the package as a whole.Possibly related prior issues: