- JMP User Community
- :
- JMP Wish List : New Ideas
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
JMP Wish List
We want to hear your ideas for improving JMP. Share them here.We want to hear your ideas for improving JMP software
We know that you have many valuable ideas on potential new features, or enhancements to existing features. You can use this Wish List to share your ideas, and discuss how to improve JMP, with other users across the globe.
Here's how to participate:
- Search: Please search for an existing idea first before submitting a new idea.
- Kudo & Comment Kudo ideas you like, and comment to add to an idea.
- Subscribe: Follow the status of ideas you like. Refer to status definitions to understand where an idea is in its lifecycle. (You are automatically subscribed to ideas you've submitted or commented on.)
- Submit: Post your new idea using the Suggest an Idea button. Please submit one actionable idea per post rather than a single post with multiple ideas.
We can’t implement everything, so please include the following to help JMP Product Management evaluate your ideas for consideration:
- What inspired this wish list request? Please describe the current issue that needs improvement or the problem to be solved that is not easy or possible right now, with an example use case.
- What is the improvement you would like to see? Please describe the idea for improving JMP. Please include mock-ups, wireframes, screenshots, scripts, other documents or examples from other software that help describe the change you would like to see.
- Why is this idea important? Please describe the value to you and/or other users if the idea is implemented (for example, ease of use, must have,…).
Status:
Acknowledged
Submitted on
01-13-2022
03:40 PM
Submitted by
dmmdiego
on
01-13-2022
03:40 PM
In the XGBoost add-in (https://community.jmp.com/t5/JMP-Add-Ins/XGBoost-Add-In-for-JMP-Pro/ta-p/319383), there is a very nice option where using DoE, a hyperparameter search can be launched, to identify the best combination of hyperparameters to maximize a given objective function.
This option is extremely helpful, and it's a standard approach when training models in R or Python.
I think there would be tremendous value implementing the same option for other predictive modeling platforms, such as Neural, Boosted Tree, Boostrap Forest and Support Vector Machine. In each case, an experimental matrix could be setup to define the search ranges for the selected hyperparameters.
The XGBoost add-in accomplishes this very nicely, so you guys already figured out the mechanics, it's a good time to make it available for other models!
... View more
See more ideas labeled with:
Right now I can do K-fold crossvalidation when building a neural network: And with the recent XGBoost add-in (https://community.jmp.com/t5/JMP-Add-Ins/XGBoost-Add-In-for-JMP-Pro/ta-p/319383), it's even better as multiple folds can be added, which are orthogonal to each other: However, this option is not available for other platforms, such as Boosted Tree, Boostrap Forest, or even a simpler Multiple Linear Regression model. Only you can add a single validation column with a unique hold-out sample. There is nothing unique about k-fold crossvalidation to XGBoost, so this option should be available for all platforms, to allow a more generalized error metrics across models consistently evaluated across multiple platforms.
... View more
See more ideas labeled with:
Right now in JMP it's very easy to select rows which are hidden or excluded. But there is no equivalent option for selecting columns which are hidden or excluded. I often use data tables with hundreds or even thousands of columns, and if I want to subset a table to only include rows and columns which are not hidden or excluded, this cannot be done (it's easy for rows but not for columns). It's very cumbersome to individually select columns when you have hundreds of them, and the "undesired" columns are already excluded or hidden. This should be easily supported through the UI, as it is a very common use case.
... View more
See more ideas labeled with:
Add Data Type icon, similarly as Modeling Type already is, to column listings. Most likely best to add as new preference option in Tables menu if some people want to keep old behaviour (but enable by default as it would already help solving issues in Community :)). This should be extended to all column listings which already have Modeling Type visible.
Example with Character and Numeric columns (of course JMP has few more)
Also quick option to change datatype could be nice, but I guess this could cause issues.
... View more
See more ideas labeled with:
JMP already has Get Punctuation Characters() and Get Whitespace Characters(), but it is missing couple of functions to get characters, for example letters and digits. Python's string library does provide these constants, which would be useful to have also in JMP: string.ascii_letters string.ascii_lowercase string.ascii_uppercase string.digits string.printable In addition to that, it might be helpful to make the functions work in such a manner, that user could pass optional argument, which would let the function to return list instead of string. Of course user can do this fairly easily (if they know how) by using Words().
... View more
See more ideas labeled with:
Hello JMP team, I'd like to suggest the following options for the partition platform: A start over button so that you don't have to press the Prune button a bunch of times to restart a tree An ability to set the minimum split size to a % of the training N qty vs the current option of just a static number The ability to stop splitting on a logworth value (for example, don't split if the splits logworth is below 2) The ability to highlight the rows in the datatable by clicking on the row of the leaf report Thanks for considering these requests! Steve
... View more
See more ideas labeled with:
Hello, Can the Group By option in the Oneway Bivariate Fit platform be enhanced to support grouping by multiple columns? Currently, it only supports one column, which can be limiting. Thank you.
... View more
See more ideas labeled with:
Hello Please find here a suggestion for improvement of scientific output 1. The model screening platform in JMP16 produces a summary table of the models across the Test set folds, with mean AUCs but no other mean metrics with would be also of interest in unbalanced data sets (e.g. F1, MCC, sensitivity, specificity, TN, TP, FP, FN, Accuracy). Indeed, unbalanced data are common in medical research were the outcme of interest is rare (e.g. disease, mortality , complications) in a data set. 2. In the drop down menu (red triangle next to model screening): "Save results Table".. generates tables for each train/validation and test set but without the above mentioned metrics (e.g MCC, ... etc), plus the metrics are not summarized as means per method (model) such as in the summary table of the models across "folds". Thanks Lu
... View more
See more ideas labeled with:
Control Chart Builder does not show an option for "Row Legend" as do other graphs such as Bivariate. Legacy Control Charts had the option to right-click and select "Row Legend".
(From Scripting Index):
Names Default To Here( 1 );
Open( "$SAMPLE_DATA/Big Class.jmp" );
biv = bivariate( y( :weight ), x( :height ) );
rbiv = biv << report;
framebox = rbiv[Frame Box( 1 )];
framebox << Row Legend( "age", color( 1 ), Marker( 1 ) );
Following this thread in the community, there appears to only be the option to make a floating legend window and then using scripting to combine. add Row Legend to Control Chart Builder output?
Since Control Chart Builder is replacing legacy control charts, would be nice to be able to add Row Legend to Control Chart Builder. One other work-around can use mark and color by row and then add Row Labels, but makes the Control Chart very busy.
... View more
See more ideas labeled with:
It would be great to see the following added into the profiler: The ability to reset the level settings back to the default setting (what the settings were when the profiler is first made, I elective this is the average of each factor) A button next to the desirability box (where you can drag the desirability line to max/min/target interactively) that you can click to run the maximize function (so you don't have to go into the red triangle) Add a summary line below all the remembered settings that would show the range of the settings for each factor. This would allow you to see which factors had the most distance between their levels for the predictions your were optimizing on. Often times, the bigger the difference between a factors levels the easier of a change it is to make so this would be great to see quickly. Be able to show all the saved profiles on the same graph. Right now this can only be accomplished by copying the frame content and pasting the frame content over the other setting (such as copying a minimize setting over a maximize setting).
... View more
See more ideas labeled with:
Hello, I often experiment with several packages installed in my local conda environments. I would like to be able to use them to create new columns in JMP data table to benefit from JMP's graph builder capabilities. More concretely, I would like to create a .JSL script that: sends the current state of the JMP data table to a python script .py or a python notebook .ipynb that is executed on a specific conda environment installed on my computer and then sends back the result of the script/notebook back to the JMP data table and updates it. Ideally, I would like to simply have to specify the python script or notebook's path and the conda environment's path in the .JSL script. For the time being, a blank "template" on how to achieve this with the current JMP version would also be very appreciated. Thank you. Best,
... View more
See more ideas labeled with:
Signal processing is important as a preprocessing process for analyzing manufacturing data such as equipment processing waveforms. "wavelet transform" will be added to JMPPro17, but at the same time I would like to add a simple signal processing "Fast Fourier transform". Various signal processing is necessary for preprocessing of time series analysis including waveforms. This signal processing can be converted into structured data, expanding the range of data that can be analyzed by JMP. In the manufacturing industry, it is very important and needs are also large. I would like to do more strongly than that.
... View more
See more ideas labeled with:
Factor analysis is the most important for quality control in the manufacturing industry.Therefore, experimental design methods and multivariable analysis are used, but these techniques assume a simple model.However, it is not appropriate to analyze complex phenomena.SEM has been added to JMP16 as a method for analyzing complex phenomena.However, SEM is very complex to operate.Therefore, I feel that the difficulty is high.As a preprocessing, we believe that a method that can analyze the data as it is necessary.For example, glasso, LiNGAM, Bayesian-Networks.I think these techniques are also useful for many companies.
... View more
See more ideas labeled with:
Allow user to select multiple subexpressions in formula editor and show them all in Preview. This would make formula creation easier. Also Allow user to make formula expressions preview table to datatable with selected subexpression and formula attached
... View more
See more ideas labeled with:
Currently when you make the preview table to datatable you will loose the formula and subexpression formula. It would be helpful to be able to export the formulas to smaller table. Also Allow selection of multiple subexpressions in formula editor
... View more
See more ideas labeled with:
Provide row number to formula editors Preview window. This could make it easier to debug formulas when you have a possible problem within subexpression. Would have to figure out how to show row numbers when using Random n, I would be fine with "real" row numbers from datatable.
... View more
See more ideas labeled with:
Currently depending on the size of data / data structure which I want to get the uniques from I use following:
Data table column values
Associative Array (if I'm lazy, but gets slow with lots of values)
Summarize (a bit better, also slows down)
Summary table and get values (I can get counts this way if I want to)
Distribution platform and get values from Report layer (nice thing with this, is that I can get other statistics at the same time and it is fast)
Lists
Associative Array
Insert values to table and use methods mentioned above
My suggestion is to add specific function to get unique values from from data tables and lists in a list/matrix. It could be also data structure like Python's Set (and if were to be data structure
A set is an unordered collection with no duplicate elements. Basic uses include membership testing and eliminating duplicate entries. Set objects also support mathematical operations like union, intersection, difference, and symmetric difference.
... View more
See more ideas labeled with:
Can you please add in a function to have a custom report that shows the changes from a prior process capability to a new process? Right now you have to cut and paste the output from two different capability reports into a single power point slide. This is how I'm used to seeing it:
... View more
See more ideas labeled with:
Status:
Yes, Stay Tuned!
Submitted on
12-14-2021
08:55 AM
Submitted by
Sizemokj
on
12-14-2021
08:55 AM
https://shap.readthedocs.io/en/latest/example_notebooks/tabular_examples/model_agnostic/Simple%20Kernel%20SHAP.html
-- Generate Python code from model - Use Model to Generate Shapley Values for local/global interpretations.
Integrating JMP PRO Python Model (NN) with explainable AI SHAP values
... View more
See more ideas labeled with:
Hello JMP, My wish: CV in Graph Builder The graph builder does not offer the coefficient of variation (CV) in the summary statistics (screenshot below, JMP 16.1). I could have used it several times so far. Everywhere else (Tabular, Summary, etc.) it can be selected, but not in the graph builder. Rob
... View more
See more ideas labeled with:
-
Data Exploration and Visualization
553 -
Automation and Scripting
443 -
Mass Customization
326 -
Data Access
236 -
Basic Data Analysis and Modeling
234 -
Data Blending and Cleanup
224 -
Content Organization
179 -
Sharing and Communicating Results
167 -
Advanced Statistical Modeling
141 -
Quality and Process Engineering
118 -
Predictive Modeling and Machine Learning
105 -
Design of Experiments
90 -
Reliability Analysis
60 -
Consumer and Market Research
43
- © 2024 JMP Statistical Discovery LLC. All Rights Reserved.
- Terms of Use
- Privacy Statement
- Contact Us