Ability to have subdaily data #132
Conversation
Merge branch 'master' of github.com:USGS-R/EGRET into subdaily Conflicts: README.md
| @@ -49,8 +49,6 @@ populateDaily <- function(rawData,qConvert,verbose = TRUE,interactive=NULL){ # | |||
| dateFrame <- populateDateColumns(rawData$dateTime) | |||
| localDaily <- cbind(localDaily, dateFrame[,-1]) | |||
|
|
|||
| localDaily$Date <- as.Date(localDaily$Date) | |||
There was a problem hiding this comment.
I was under the impression you were only going to change the plotting functions, and if people used EGRET "as is" they wouldn't notice a change. Are you certain that deleting this line assures 99% of EGRET users will still get Date objects?
| @@ -146,6 +146,60 @@ test_that("modelEstimation window params work", { | |||
|
|
|||
| }) | |||
|
|
|||
| test_that("modelEstimation works for POSIXct and Date classes", { | |||
There was a problem hiding this comment.
I don't think we can say modelEstimation will work with sub-daily timesteps. There are a bunch of steps in the modelEstimation code that assume daily time steps. We'd have to very very carefully go through all of that and re-do some logic for sure (we assume 1 row equals 1 day, we can't do that anymore). I would honestly just take these tests out because I think it would only confuse future us.
|
Closing and resubmitting just plot label changes per Slack conversation about how I probably got carried away with dates :) |
This PR removes the code that forces dates to be class Date, and allows POSIXct. This is desired by
loadflex, see DOI-USGS/loadflex#135