Refactor 3: Behavior - Githubissues

During the harmonization refactor, I noticed a few functions could use some minor fixes/improvements that will (or could potentially) change their behavior or how they are used:

FUNCTION REFACTORS:

getDaylightHours

[x] Evaluate bbox default and documentation. Despite the param doc stating that it defaults to CONUS, the param is set to null. The function is fairly location specific--so it's probably better that bbox defaults to null. I'll change the docs.
[x] Remove @rdname tag in the documentation.

goesaodc_areaPlot

[x] ... param is documented as "Additional arguments passed to \code{goesaodc_plotSpatialPoints()}." but is never actually passed. Fix this so it is!
[x] If nc is a single ncdf4 handle, put it into a single element list (like goesaodc_createNativeGrid()).

goesaodc_createNativeGrid

[x] verbose param is never used. Not sure what information should be printed out here...maybe it wasn't supposed to have the param in the first place? I see this function calls ncdf4::ncvar_get(..., verbose=FALSE, ...) several times, so I suppose I could replace those with ncdf4::ncvar_get(..., verbose=verbose, ...), but that spits out a crazy amount of of information which doesn't even fit in the R console. I think I will get rid of the verbose param for now.

goesaodc_createRasterStack

[x] Verbose printing should include an actual measure of progress like stacked layer (12/58): ....
[x] Default bbox to the bbox_CONUS package variable instead of defining its own vector.
[x] Remove @rdname tag in the documentation.

goesaodc_createSpatialPoints

[x] Default bbox to the bbox_CONUS package variable instead of defining its own vector.

goesaodc_downloaddaytimeAOD

[x] Evaluate bbox default and documentation. Despite the param doc stating that it defaults to CONUS, the param is set to null. I think it makes since for the param to default to null, so I'll update the docs.
[x] Remove @rdname tag in the documentation.

goesaodc_listDaytimeFiles

[x] Evaluate bbox default and documentation. Despite the param doc stating that it defaults to CONUS, the param is set to null. I think it makes since for the param to default to null, so I'll update the docs.
[x] Include a baseUrl param that is passed to the internal goesaodc_listFiles() call.
[x] Remove @rdname tag in the documentation.

goesaodc_listFiles

[x] Force users to specify hour for datetime and endtime.
[x] Error message when enddate is before startdate.
[x] endtime should be the first time excluded from the end of the time range.

goesaodc_plotSpatialPoints

[x] Add param validation.

raster_createLocationTimeseries

[x] Add param validation.

raster_createStackAverage

[x] Move into utils_raster.R.

utils_goesaodc.R

[x] Add param validation to all functions.

GENERAL REFACTORS:

Validation of datetime types

[x] goesaodc_createRasterStack() and goesaodc_createDaytimeRasterStack() validate datetime/enddtime with:
```
# Validate if time passed in is already a posix time with timezone
time_classes <- c("POSIXct", "POSIXt", "POSIXlt")
if ( class(datetime)[1] %in% time_classes ) {
timezone <- attr(datetime, "tzone")
}
```
while goesaodc_ListFiles() uses:
```
# Use timezone from POSIXct datetime, or passed in timezone
if ( lubridate::is.POSIXt(datetime) ) {
timezone <- lubridate::tz(datetime)
} else {
if ( !timezone %in% OlsonNames() ) {
stop(sprintf("timezone \"%s\" is not recognized", timezone))
}
}
```
Are they the same? Is one preferable to the other? I like the lubridate version for its readability, but lubridate is only used 5 times throughout our package and it seems like overkill to import if we use it so little. Since the lubridate version seems more clean and "safe" though, I'm going to switch functions to use that for now.

Verbose display

[x] Any functions that print verbose progress messages use message() instead of print().

MazamaScience / MazamaSatelliteUtils

Refactor 3: Behavior #79