Rename classes #751
Rename classes #751
Conversation
Including a couple of small fixes that are unrelated and needed to be done, but didn't result in errors so hadn't been caught by the CI: Changing the `workflow` kwarg to `parent` in a node instantiation, and updating the description of how things are (or aren't) automatically updated in cell 17.
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
|
|
# Conflicts: # pyiron_contrib/workflow/is_nodal.py
It's not exhaustive or finalized, but in general the universal stuff now sits on `Node` and the stuff specific to wrapping functions sits on `Function(Node)`
And have them fail readiness when this is their value
When building input channels in function nodes. And test to make sure it stays this way!
Maybe the implementation test should even be removed...
Now that Function is "fast" by default and we don't explode when no defaults are specified
|
Is there a possibility to run tests here? It looks fine but without tests it's difficult to say... |
We could rebase this onto main briefly, which will trigger tests on this (including everything this is stacked on top of), or we can change the centralized CI as you and I propose to do/discuss in the Monday meeting. For each of this stack of PRs I made sure the workflow tests are passing on my local machine; it will still be necessary to do the full CI on other architectures, but I guess it's already quite promising. Not running it here just delays catching problems until we get down to the main-most of the stacked PRs, but we'll still catch them with our current workflow. |
To avoid cluttering the tab completion menu
Fast by default
Data channel value starts as not data
Since the tests are running on my machine and we're not merging into |
Rename
NodetoFunctionand thenIsNodaltoNode, including downstream changes like changing the decorator from@nodeto@function_node, the childrenFastNodeto justFast, etc.In the process of updating the example notebook, I also made two small fixes to problems that had slipped through the cracks because they didn't raise any errors:
workflow=wfkwarg toparent=wfin one of the instantiationsNot included in this PR: Moving all the docstrings around so that they better conform to reality and the spec in #743 -- right now
Function(Node)has a bunch of pretty general specifications that should probably be moved over toNode, and the idea of macro nodes is not yet mentioned.