I have a datatset with secondly values. The dataconsitency is not so good so sometimes there are also days missing. I want to get the mean value of every hour in this dataset. The time should also contain the date. So that I get something like this:
2015-01-01T00:00:00', '2015-01-01T01:00:00', '2015-05-01T02:00:00'....... and the mean variables to it.
I tried:
ds_mean = ds.groupby("time.hour").mean()
But there it gives me a mean value of all hours a this time containing the dataset.
Then I tried to create a dataset for each day. But the problem is that the time only contains the hour.
Instead of groupby, use ds.resample
. Groupby groups on the unique hour values (1-24) which appear in the data, whereas resample explicitly changes the frequency of a DateTimeIndex dimension:
ds_mean = ds.resample(time="H").mean()
See the user guide section on working with time series data for more info and examples.