Hello all, I am trying to run the Age prediction using machine learning tutorial and in the Analyze RNA seq data section under optimize hyperparameters in the Hands On: Hyperparameter search it says to set
param-repeat “2: Parameter settings for search”
“Choose a parameter name (with current value)”: elasticnet__normalize: False
However this option doesn’t show up and seems to be depreciated. I can run it without that option but then the output looks completely different. Is there an equivalent option?
Hyperparameter Search performs hyperparameter optimization using various SearchCVs (Galaxy Version 1.0.11.1) (link to tool)
If you are working somewhere else, and the server is listed under the Available at these Galaxies menu (at the top of the tutorial) then this same version of the tool should be available. You might need to navigate with the Versions menu. → FAQ: Changing the tool version
Please let us know if you are able to solve this! If not, you can share the URL for the server and a link to the exact step you are at, and we can look into this more!
And .. I just reread your question. Are you asking about how to expand the form to layer in additional parameters? If so, then try clicking on the button to add in another repeating block to the form.
Screenshot where I added one more repeating block, and am hovering over the button that can add in more!
Many tool forms can do this to save space with arbitrarily repeating segments. Very understandable that the first time through it seems confusing! Still, let us know your thoughts and screenshots back might help to explain if my guess wasn’t correct!
I am specifically talking about this option of elasticnet__normalize: False which does not appear to be a parameter name I can choose in the drop down.
The available parameters to filter on are based on the input dataset’s content. These will vary depending on the data. Your input contains something different from the example data, yes? If true, then you can only filter on the values already listed. If you are using example data and things look off at this step, it means that something went wrong in an earlier step.
The tutorial includes a few workflows! To see how the flow through the options work, you can compare the parameters versus the example data. The workflow invocation links have the baseline workflow, then the invocation and associated history with those inputs.
Part of this process is defining which content in the original file to consider and to preprocess it. After, that prepared content is available to the filters when using Hyperparameter.
Does this help now? If not enough and you are still stuck, you are welcome to share your history for more specific help and a review to get you back on track! → FAQ: Sharing your History