nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

346247_en-US

liliconcom

Replacing missing data by mean/regressed values

We have a add-in perform imputation for missing values. I think this needs to be embedded in JMP.Imputation Addin

Also for a time series data, It would be helpful to have feature replacing the missing values by a linear model derived from the closest available values in time. Replacing missing data by regressed values

Other

Re: Replacing missing data by mean/regressed values

Hi Mia,

Thank you for your response.

I haven't tried the methods in the Explore Missing Values platform, particularly ADI. If I found any issues, I'll comment here.

Thank you for adding the request for imputation in time series to defect.

Regards,

Yuichi

Re: Replacing missing data by mean/regressed values - Status changed to: Open for Voting

Thank you for posting this Yuichi.

For the first question, @jacob_rhyne has provided this response:

We should and would like to direct customers to use the methods in the Explore Missing Values platform. Also, in his dissertation, the developer (Milo) showed that ADI (Automated Data Imputation) outperforms mean/mode imputation, therefore I wouldn’t recommend JMP endorsing mean/mode imputation by putting it directly into the product.

Have you tried the methods in the Explore Missing Values platform, particularly ADI? If so, are there any particular issues found that would cause a user to prefer the methods in the add-in over the methods in the Explore Missing Values platform?

The request for imputation in time series has been added to this defect: http://sww.sas.com/ds/DefectsSearch/S1582/S1582644.html