write jobflow tutorial #18
Conversation
Merge Janine's addition to my own fork
|
@QuantumChemist Thank you so much!! |
jobflow.ipynb
Outdated
| "outputs": [], | ||
| "source": [ | ||
| "graph = to_mermaid(qe_flow, show_flow_boxes=True)\n", |
There was a problem hiding this comment.
Could someone who has QE set up, run these lines please?
There was a problem hiding this comment.
You can simply install it via conda. I have done this as well
There was a problem hiding this comment.
Oh cool, didn't know this 😁
There was a problem hiding this comment.
We have a wide range of opensource codes available on conda-forge. You can find the full list here - at least the packages I maintain https://github.com/jan-janssen/conda-forge-contribution . If you are interested to add any other code I am happy to help.
|
I have finished the tutorial @JaGeo 😀 |
|
@QuantumChemist I will check it out next week and will contribute as well! |
|
I can't get QE to take the right pseudopot path. I will try again on another day ^^ |
|
Somehow my notifications were not up to date and I missed this. To find the pseudo potentials in the Github Continuous integration environment we set the following environment variable Unfortunately the Github Action is not executed in the home directory, so the |
Oohh I might try to specify the QE pp env variable like this next time! And I have to check how jobflow handles env variables in general. |
adis_tools/schemas/qes_230310.xsd
Outdated
| @@ -22,7 +22,7 @@ datecode 220603 | |||
| <element type="qes:outputType" name="output" minOccurs="0" /> | |||
| <element type="qes:cpstatusType" name="STATUS" minOccurs="0"/> | |||
| <element type="qes:cptimestepsType" name="TIMESTEPS" minOccurs="0"/> | |||
| <element type="qes:statusType" name="exit_status" minOccurs="0" maxOccurs="1"/> | |||
| <element type="qes:statusType" name="status" minOccurs="0" maxOccurs="1"/> | |||
There was a problem hiding this comment.
I had to adjust quite a few things with the xml format. Maybe something is different in the xmlschema version I use?
There was a problem hiding this comment.
ofc now some tests fail... 🙄
There was a problem hiding this comment.
I removed the modifications to the parser in #19 and everything works fine. In addition, when I execute the notebook here locally I also get an xml related Reason: Unexpected child with tag 'fcp' at position 18. error, but the notebook execution does not fail. It seems like jobflow is converting the error to a warning and continues executing the remaining cells of the notebook. You can test it with
There was a problem hiding this comment.
Did you test it locally with a different quantum espresso version? In the repository we currently use qe=7.2.
There was a problem hiding this comment.
I think it's the qe version as I used qe 6.7. Thank you for looking into this and fixing this issue 😄
I will merge your commits back into mine to have passing tests in my repo.
There was a problem hiding this comment.
I also realized that I overlooked the environment.yml file 🙈
adis_tools/schemas/qes_230310.xsd
Outdated
| @@ -150,8 +150,8 @@ datecode 220603 | |||
| <element type="double" name="press_conv_thr" default="5e-1"/> | |||
| <element type="qes:lowhighType" name="verbosity" default="low"/> | |||
| <element type="positiveInteger" name="print_every"/> | |||
| <element type="boolean" name="fcp"/> | |||
| <element type="boolean" name="rism"/> | |||
| <!-- <element type="boolean" name="fcp" default="false"/> | |||
There was a problem hiding this comment.
Removing this line causes the pyiron tests to fail with:
Reason: Unexpected child with tag 'fcp' at position 18.
To me there seems to be an issue with the input files.
|
@jan-janssen I will give it another read this week. Alternatively, we can merge this for now and I make a potential new pull request with improvements. (ToDo for tomorrow) |
Both is fine for me, I still was not able to reach @mbercx so far ... |
in https://github.com/materialdigital/ADIS2023/blob/f33ebb84921d1cd907b60d3c66494eddfd2512a4/adis_tools/parsers.py#L15 for `XMLSchema` we have `Expected type 'str | bytes | Path | IO[str] | IO[bytes] | XMLResource | Element | ElementTree | list[str | bytes | Path | IO[str] | IO[bytes] | XMLResource | Element | ElementTree]', got 'Traversable' instead`
I'd say the second option is easier. Then we don't have to merge/fork back and forth 😆 |
|
Then, I will merge! |
|
@QuantumChemist just saw that there is still old text on the structure in here. We use MSONatoms instead of Structure. I will correct it |
Oh that makes sense. I was a bit confused about it, but somehow didn't further think about it 🙈 |
I started with writing the
jobflowtutorialTo-Dos:
atomate2,custodianetc.