Antonio

Thanks, both of you. I have tested @felixcremer s idea, but got a weird structure for the data. So I've also tested @Balinus s idea, but It seems that YAXArrays...

I have tried open_dataset and it works! I do not know the difference between Cube() and open_dataset(). But I need to open a lot of files into a singe dataset,...

Wow!, I understand now. I have two shared axes: "values" and "valid_time" and three YAXArrays: "longitude", "latitude" and "msl". So it is a data set, but not a Cube. On...

Here are two of them. The first one has the `msl:coordinates = "longitude latitude" ;` attribute on the msl array and the second one, not. They are in a zip...

Ok. The file is too large, so I send it by we transfer: [https://we.tl/t-AC7DBQp7vE](https://we.tl/t-AC7DBQp7vE) Here is the output of my datset, "ds": ``` julia> ds YAXArray Dataset Shared Axes: ↓...

Yes, they are converted from Grib. I can open the gribs with GRIBDatasets with no problem. In fact, I converted the grib files to netcdf by opeing the gribs in...

I have to correct myself: Raters correctly interprets the structure. I tested it a long time ago and it seem not to do it, but now it does!. So, I...

I have the answer to my question. In order to select the variable msl depending on the values of latitude (which is not a dimension, but another variable that shares...

Wow! It runs now instantaneously. But, due to my particular data structure, I have to read the variables msl and longitude in different variables: ``` ds_m = readcubedata(ds.msl) ds_l =...

But, is there a way to load a dataset into memory, other then `readcubedata`, or only `readcubedata` can load into memory?