Conversation
…terpolation nor a selector caused an error because the index_type was not respected. Solution was to apply a temporary nearest neighbor selector on the expanded coordinates.
…source._get_window_coords function
…._eval interfered with no-intersection behavior. This attempts to correct that
…did returned coordinates/indices from the input and not the source. Added override for Datasource.create_output_array that adds in bounds and boundary_data to attrs. This already happened on successful evals but short-circuit cases did not contain these attrs.
…eration is applicable to account for podpac4's separation of nodes and interpolation
…_node to separate value and units/label. expanded enumeration label behavior to handle array returns instead of just single values.
This reverts commit 4c9bf79.
…feature/general_bug_fixes
…fixed boundary_data value, added clarifying comments.
…portance to nodes doc
mpu-creare
left a comment
mpu-creare
left a comment
There was a problem hiding this comment.
A few minor comments.
Major thing as you already guessed is the lack of code coverage. Please add some tests that cover the new code.
Co-authored-by: mpu-creare <mpu-creare@users.noreply.github.com>
…to explain utils._get_label
…feature/general_bug_fixes
|
@mpu-creare When adding unit tests to improve coverage, I noticed that this catch doesn't have a well defined return. Should we return |
Very yes. Nice catch. |
Roger that. I went ahead and committed that change already (it made the unit test easier to write). |
|
3 participants
This branch contains various behavior fixes/adjustments that were discovered when working through the transition from podpac3 to podpac4 with another program that makes heavy use of podpac