Feature/increment write #15
Conversation
…ture/increment_write
…er introspection of file contents. There seems to be issues around time/tile dimension and data, currently commented out.
…es to output increment files, and other tweaks.
| if (fieldMetadata.lfricWriteName != "TO BE DERIVED" && | ||
| fieldMetadata.lfricWriteName != "TO BE IMPLEMENTED") { | ||
| if (fieldMetadata.copyFirstLevel == true) { | ||
| atlas::util::Config atlasOptions = atlas::option::name(fieldMetadata.lfricWriteName) | | ||
| if (writeName != consts::kToBeDerived && writeName != consts::kToBeImplemented) { |
| int getDataSize(const atlas::Field& field) { | ||
| std::vector<int> dimVec = field.shape(); | ||
| int size = 1; | ||
| for (const auto& dim : dimVec) { | ||
| size *= dim; | ||
| } | ||
| return size; | ||
| } |
There was a problem hiding this comment.
Think this includes halo points in the count - is this what you want? Is it for global fields only?
There was a problem hiding this comment.
This is a very good spot. I didn't want the ghost points, I don't think. Let me play with this. There'll be an update on this, I'm sure. Thank you!
There was a problem hiding this comment.
Ah yeah. I checked, @mo-joshuacolclough. I am using it with global fields only. So the current implementation is what I need. Glad you made me think about that though!
I'll rename the function to make it clear. Otherwise I'll probably forget later on!
|
Thanks @mo-joshuacolclough. The change to |
This PR introduces the necessary changes to MONIO for use of
Monio::writeIncrementsfrom LFRic-Lite.Current test outputs: http://fcm1/cylc-review/taskjobs/punderwo/?suite=monio_increment_write_05
The work in MONIO for use of
Monio::readIncrementswill come next, but the work in LFRic-Lite to use both of these functions will appear in a single PR.