JMP Wish List

Sandeep123 · ‎04-02-2021

Certain very simplistic things are missing in the JMp platform (1) JMP cannot do propensity score matching (2) JMP does not calculate Cramer's V or Cohan's D for effect size calculation (3) JMP cannot do multiple imputation for categorical variables (only does informative missing option) (4) JMP does not allow random effects in logistic regression models I may be mistaken in some of these. But I have asked this question in the discussion forums and as far as I can tell are accurate. Some of these are real bummer's

reese_sj · ‎02-08-2021

It would be useful to be able to recode column names, but only ones that are selected. Often with larger data sets there is only a subset of columns that need to be modified or there is a variety of naming conventions used. In these cases being able to recode a subset would allow for more customization with the recode tool.

mark_anawis · ‎01-14-2021

It would make scripts and analysis run faster if we could work with rows in similar fashion to columns. The current workaround is to stack or capture columns as a matrix, transpose, and put back into data table so row 1 becomes Col1. This take time and memory for large data tables. If we could use a row of data (or matrix) in place of a column(:Col1) below to generate a graph, that would be very useful. In this example, I am creating graphs to save into a column using a loop for each row (not shown). I first had to capture the columns I want graphed into a matrix, transpose it and put it back into a new data table (dt_mat). Add a new column ("Read_Number") which is 1,2,3,4,5..... The original data was in columns: :read1, :read2, :read3, :read4, :read5,...... After transposition, it becomes :Col1, :Col2, :Col3, :Col4, :Col5, .... where the columns are same as untransposed rows from original table. The pic object is inserted into the table as part of loop. The actual code uses a list of column names in place of :Col1 below. dist = dt_mat << Bivariate( Y( :Col1 ), X( :Read_Number ), Fit Each Value( {Line Color( {212, 73, 88} )} ) ); pic = (Report( dist )[Outline Box( "Bivariate Fit of Col1 By Read" )]) << getpicture;

shampton82 · ‎01-12-2021

I'd like to have the ability for the minimum split size for tree models to be based upon a % of the training sample size vs a integer number. So for example I could ask that the minimum split size would be 10% of the training samples size, rounded up.

HubP_SDe · ‎12-24-2020

Dear Team, If I'm not missing anything, the App Builder objects Tree-View could be featured with following improvements: Why is it apparently not possible to run an App from the Script Tab? Always getting the error below if not launched from a Module Tab: Name Unresolved: OnModuleLoad in access or evaluation of 'OnModuleLoad' , OnModuleLoad( {} ) /*###*/ Ability to move items around in hierarchy, to drag one or more items in/out whatever container. Btw, just experienced JMP immediate shut-down at double-cliking a tre-view object from the Script Editor tab. Thanks!

MathStatChem · ‎12-09-2020

It would be very nice if Evaluate Design allowed you to use a local data filter when evaluate a single design (single table). The use cases are: Evaluating the impact of lost design point(s) Quickly seeing evaluation of intiial and then subsequent augmented designs. I'm not suggesting you should be able to compare designs this way, this is only for rapid check of design properties for subsets of a full design table.

David_Burnham · ‎12-05-2020

Applying a split to the JSL editor window causes the dirty flag to be set (i.e.the save icon becomes enabled). Please don't do this! Reserve the dirty flag to code changes only. Use - Case: Seeing the save icon appear on an code that I am reviewing causes me to immediately close the window in case I have accidentally interfered with the code - you know the type of thing, where you have typed 'c' into the code when you meant to do CTRL-C).

jsdejmp · ‎11-18-2020

It would be helpful to have a way for me to record structured notes as I progress through an analysis. It can be very critical to track why an analysis was done a certain way or why an outlier was removed. My thinking is to be able to create - in the same place Scripts are stored - a new type called a Note that features a multiline rich-text field for users to add instructions, comments, etc. (I tried adding only comments to a new script using /* and */ and the script wont save unless there is a non-comment element in the script, even "", so I feel like this would be a low hanging fix - allow scripts to save whatever is typed, even if its only a comment block)

jambo · ‎11-16-2020

Hi Would it be possible to get a native propensity score matching app in jmp like we have in SAS . It could be quite helpful for us clinicians . thanks sajid

kevin_herlihy_l · ‎11-03-2020

Dear JMP Team, I'd love to have the ability to generate a summary table where the data itself was linked to the source table. And if I'm being greedy, I'd like to be able to use the summary table to generate a table of stacked columns which also link back to the original source table. This would ultimately result in the ability to change the source data and have the change reflected in a stacked summary table. Thanks, Kevin Herlihy

DeanB1066 · ‎10-12-2020

JMP platforms have inconsistent defaults. For some, you can set a default confidence level, e.g. the Distribution platform. Matched pairs has customizable default, but it is called "Set [alpha] level". We use a standard of confidence = 0.8. We can set that default for all the various platforms we use, except for "Life Distribution". I may be the only person submitting this change, but it is a standard within all of Department of Defense testing.

ThomasZatValcon · ‎09-04-2020

I would like a "Stack properties" checkbox added to the Stack command. The result should be that each property in the stacked columns (where feasible) creates a column with that property name. Each row of these property columns contain the original properties of the stacked columns. This would make the stack command more complete, as currently most of the properties are lost when stacking columns. This ways the properties could still be made available when going from wide to narrow format. The split command could have similar (opposite) functionality, i.e. a number of columns could be added to a property columns list, and when splitting the contents are added to the properties of the created columns.

nathan-clark · ‎07-14-2020

I would like to see the ability to select cells by their color, not value. This comes as I have learned of the imputation ability for missing values and JMP will color imputed cells the same color. If the window is closed and 'undo' is no longer available, it would be great to be able to select those cells by color to exclude/undo/etc. (another way would be to update imputation to make selection easier after the fact, but the suggestion I believe would have use beyond the noted use case).

txnelson · ‎06-18-2020

For character columns, the Distribution Platform has an option to Test Probabilities. The Hypoth Prob column that is displayed and manually entered, needs to have an option to split the Probabilities equally among the categories displayed. This option is a very common selection and would make the usage of this analysis a much friendlier tool.

Hoeppner · ‎06-16-2020

Match JMP probit analysis with PROC PROBIT from SAS, including LACKFIT option. SAS will assume normal or T distribution depending on the lack of fit test result. LoD and CI results from the two platforms can vary based on this difference.

MathStatChem · ‎05-19-2020

When you try to recode a column that has the List Check property, and the recoded value you want to use is not in the list of items for the list check, Recode gives you a warning that this has happened, but then it appears that it will allow you to do the recode. However, when you click ok the recoding does not actually happen. I think it would be good to have the option in Recode to either delete the List Check property or update the list check property with the new value that has been added with recode. The use case is when you have a column with a list check that has missing values, and you want to recode the missing values to another value (e.g. "Missing") so that you can use that as an informative missing category for prediction purposes.

Thierry_S · ‎04-29-2020

Hi JMP Team, Would it be possible to enhance the Column Reorder command by allowing sorting of only subset of selected columns? Thanks, TS

bemevar · ‎04-28-2020

I wish JMP has "Econometrics and Time Series Analysis" capability/functionality. This is quite basic, even SPSS has it. Or, maybe I am missing something?

txnelson · ‎03-07-2020

I would like to right mouse click on a rowstate column and have an option to select matching "colors" or "Markers" or "Colors and Markers"

Jim_Gillard · ‎02-14-2020

Would like to have the ability to actually hide rows that are selected to be hidden. Makes review of specific rows quick. Otherwise will just keep using Excel.

JMP Wish List

Missing statistical features

Recode Column Names- Only selected columns

Using rows to generate graphs

Minimum split size on trees

App Builder Tree-View improvement

Add the ability to use a local data filter in Evaluate Design

Script Editor Dirty Flag

Add "Notes" data type to scripts window

Propensity score matching

Virtually linked summary tables and stacked columns

Customize Default for Confidence Level in Life Distribution platform

"Stack properties" option in the stack command

Cell Selection by color

The Chi Square , Distribution Platform, Test Probabilities, a selection to set all Hypoth Prob to = Prob

Probit update

Recoding Column with List Check Property

Column reorder > Sort by Name / Data Type / Role > SELECTED COLUMNS ONLY

Econometrics and Time Series Analysis

Allow for "Select Matching Cells" for the rowstate column

Row Filtering and hiding