Where start to learn JMP 101 for data cleanup and visualization.

[#1]

I see a lot of learning resources and various topics to the point that it is overwhelming, so maybe someone will be able to help me out. I have a large dataset from REDCap of about 10k people with 300-400 variables per line. I was wondering what would be the most effective way to start analyzing it. At this moment I'm not really interested in any statistical analysis, mostly just to clean it up and visualize, once that is complete, I plan to look into to clinical data and treatment outcomes.

Thank you up front!

[#2 · Accepted solution]

Hi @MathProductPuma ,

I think the best way to start in your situation is here: Learning Hub / JMP Community .

You can either go for short solutions (Minute Guides; ordered by application area) or to all the materials for a specific application area, e.g. "Data Blending & Cleaning". Simply click the tile in the workflow graph.

I thinks that's as specific I can go here, unless you have more specific descriptions of tasks you need to solve.

Best

Florian

[#3]

thank you! this seems like a right place to start!

[#4]

Also check out these tutorials to see if anything fits your needs:

https://community.jmp.com/t5/Mastering-JMP/tkb-p/learn-jmp-tutorials-tkb

[#5]

thank you! looks useful

[#6]

Opposite from your plan, I suggest that you start with a statistical analysis objective, then try to find how to address that objective with JMP. Your data-cleaning and data-visualization are statistical analysis objectives in nature. Why do you think your data is not clean, any evidence, how do you verify your suspicion? Why do you want to visualize your data, for what, why?

[#7]

thank you for the suggestion. I have looked at it and see that not all forms are filled out, some have incorrect entries and not all of them are equally useful. cleaning it would let me prioritize which ones would I like to look at.

[#8]

What you have described lead to following thought:

• "Not all forms are filled out"
  Do you mean entire forms or some entries in the form? If you mean entire forms, then I think that is a survey specific subject, because you don't have all responses. You need to search survey analysis field to find solutions. If you mean missing entries, then it is a general subject called "missing values". There are different approaches: e.g. exclude observations from analysis (I bet this is not ideal to you); treat missing as a separate category; impute, etc. All are what you can search to see how to do in JMP, and try to understand their consequence to your downstream analysis.
• "Not all of them are equally useful"
  Do you mean entire field across all responses? Then you need to decide whether to keep it or not in your analysis. This is a subject probably under the name "variable selection". Search it and see what you can do.
  Or do you mean some entries are not useful? I guess, then you can decide whether to replace those entries with missing values.