Open mmoda-esg-bot opened 3 months ago
Have you tried the table_conoute tool? It's a pandas wrapper with a lot of functionality
Thanks! We tried a few but did not manage to get what we need, not easily anyway. What we want is something like c1, c2*sin(c1)
edit: also we need hist2d
equivalent, is there one?
Im.on my phone but have you tried https://usegalaxy.eu/tool_runner?tool_id=toolshed.g2.bx.psu.edu%2Frepos%2Fiuc%2Ftable_compute%2Ftable_compute%2F1.2.4%2Bgalaxy0
We did and it we did not manage to compute c2*sin(c1)
for some reason.
@bgruening please let us know if we missed something!
Here is my first attempt: https://github.com/galaxyproject/tools-iuc/pull/6258
Here is my first attempt: galaxyproject/tools-iuc#6258
Thanks! We thought about proposing this sort of change, but we wanted to complete the workflow quickly so our choice was clearly faster. We anticipated that if we find another way (e.g. with your advice) we'll adopt an existing common tool.
By the way, I wonder if the tool is restrictive to avoid running arbitrary code like we do in ours here?
By the way, I wonder if the tool is restrictive to avoid running arbitrary code like we do in ours here?
Yes, that's the point why we need to have an allowlist
and only allow certain actions. I'm very hesitant to deploy anything on EU that can run arbitrary code. The ITs already allow a lot of theses and people abuse them :(
By the way, I wonder if the tool is restrictive to avoid running arbitrary code like we do in ours here?
Yes, that's the point why we need to have an
allowlist
and only allow certain actions. I'm very hesitant to deploy anything on EU that can run arbitrary code. The ITs already allow a lot of theses and people abuse them :(
Interactive jupyter runs arbitrary code on EU instance.
Do you hesitate because of the security implications?
Or because it might make the workflow harder to analyze and understand?
Do you hesitate because of the security implications?
For sure.
Or because it might make the workflow harder to analyze and understand?
A bit.
Interactive jupyter runs arbitrary code on EU instance.
That is what I meant with IT
. And they are super restricted and still cause problems. We need to have something like this here in place: https://github.com/usegalaxy-eu/WallE/
Do you hesitate because of the security implications?
For sure.
It does create some tension with ease of workflow development. I suppose if tool like ours has some the parameter that defines the code "frozen" - i.e. not changing in the workflow, it could be certified as safe? I.e. workflow can be safe even if the tool is not. Is there a mechanism for something like that?
Interactive jupyter runs arbitrary code on EU instance.
That is what I meant with
IT
. And they are super restricted and still cause problems. We need to have something like this here in place: https://github.com/usegalaxy-eu/WallE/
Ah I was thinking about other IT and was wandering what it implies. How are they restricted, not preventing some code, right? Any code can be run?
btw. the new tool is now deployed and available: https://usegalaxy.eu/root?tool_id=toolshed.g2.bx.psu.edu/repos/devteam/column_maker/Add_a_column1/2.1
btw. the new tool is now deployed and available: https://usegalaxy.eu/root?tool_id=toolshed.g2.bx.psu.edu/repos/devteam/column_maker/Add_a_column1/2.1
Thanks!
Oh we found generic awk tool already deployed, seems to work?
Yes. Awk is available since years. But specifically implemented a sandbox mode for awk back then.
Yes. Awk is available since years. But specifically implemented a sandbox mode for awk back then.
I think we can implement lots of things in awk, even in sandbox. Certainly it has cos etc. Maybe we can wrap it in subworkflows for frequent operations.
Have you tried the table_conoute tool? It's a pandas wrapper with a lot of functionality