This release fixes the NASA POWER requests for sf objects and several
indices. Some results change. Check the items under CHANGES IN RESULTS
before comparing with analyses run on earlier versions.
rainfall() computes SDII as the rain in wet days (>= 1 mm) divided by
the number of wet days, as in Aguilar et al. (2005). Earlier versions
divided the total rain, including days below 1 mm, by the number of wet
days, which overestimated SDII.temperature() returns DTR (diurnal temperature range) as a decimal
number. Earlier versions truncated it to an integer when
timeseries = FALSE.crop_sensitive() reports hts_*, hse_* and lethal_* as the
proportion of days with data. Earlier versions divided by all days,
including days with missing data. The documentation now says proportion
(0 to 1) instead of percentage.crop_sensitive() assigns cdi_mean to the daily mean temperature and
cdi_max to the daily maximum temperature (#19, #20). The two were
swapped before.array methods of GDD(), ETo() and late_frost() include the
last day of the data when neither span nor last.day is given. Earlier
versions dropped it.temperature(), GDD(), ETo(),
crop_sensitive() and late_frost() for sf objects, failed before.
Values are matched to dates and grid cells by name, so a parameter whose
name is part of another (for example T2M and T2M_MAX) is no longer
mixed up. Missing values coded as -999 are returned as NA.rainfall() for sf objects uses "PRECTOTCORR" by default. The previous
default, "PRECTOT", is no longer offered by NASA POWER.sf objects are read with 'sf' by namespace, so 'sf' no
longer needs to be attached. Polygon centroids are computed on the
geometry only; other columns of the object were mixed into the
coordinates before.GDD() with equation = "c" stops with a clear message when
tbase_max is missing. The documentation of equation "c" now says that
tmax is capped at tbase_max.rainfall() and temperature() with timeseries = TRUE stop with a
clear message when intervals is missing or longer than the series.cumrain() and cumdrought() return a vector of the same length as the
input when it has a single value..st_span() requests NASA POWER data up to the last day actually needed
when sites have different start dates and spans.GDD(), of the
hse and hse_ms indices, and of the *.threshold arguments. The
Overview vignette passes tbase to late_frost() (it passed base,
which was ignored). The README example uses the right dates column.R CMD check (they were excluded in
.Rbuildignore), with new known-answer tests that do not need the
internet.AGENTS.md, a guide for AI agents working on the package.temperature.data.frame() and ETo.data.frame() in requesting data from nasapower as the API cannot handle more than one parameter (TMAX, TMIN)GDD()GDD() and late_frost()crop_sensitive() to compute crop sensitive indices.late_frost() to compute late spring frost.get_timeseries()data.from in get_timeseries()last.day, optional to span, when indices are retrieved with focus on time series analysisstats::hclust()print() method for objects that inherits the class 'clima_df'pars, mainly for temperature(), GDD() and ETo(). Details in the function documentationtemperature() indices provided. Details in the function documentationGDD() applies two equations adjustments for cold zones and one for warm zones with the argument equation. Details in the function documentationGDD() now allows to return the raw daily gdd or number of days to reach the accumulated gdd (the default in previous versions) using the argument return.asGDD() arguments. Previous versions used GDD(object, day.one = NULL, degree.days = NULL, base = 10, span = 150, ...). From now on GDD() default behaviour is GDD(object, ..., base = 10), with additional arguments are passed through ... in each defined method. The function return by default the accumulated degree-days across the series.modis() and chirp()R to >= v3.5.0