JMP Wish List

Would like to be able to generate random values from a Weibull distribution where the threshold parameter is not 0. ... View more

What inspired this wish list request ? Scientific publications I read about the proposed improvement   What is the improvement you would like to see ?    Addition of options to perform T² multivariate control chart with MCD and MVE estimation method (Minimum Covariance Determinant, Minimum Volume Ellipsoid resp.).  Details about these methods are provided in attachment.   Why is this idea important?  T² multivariate control charts are based on the estimation of a mean vector and a covariance matrix. These estimations are sensitive to the presence of outliers in the dataset and can lead to a masking effect preventing from detecting some outliers. MCD and MVE are so called "Robust" estimation methods, much less sensitive to the presence of outliers. It significantly improves the detection level of multivariate outliers.   ... View more

What inspired this wish list request? These two community postsColor code Top 5 and How to use JSL to set the color gradient of a column in text format by the value of a numeric column?. Interactive option can be a hassle (and not sure if it works with gradient) (and JMP couldn't capture those actions for automatization) and other option is to go with scripting.   What is the improvement you would like to see?  I would like to have some option to color cells based on other columns colors. This could maybe be implemented as column property?   Why is this idea important? Improves the visuals of JMP datatables .     ... View more

What inspired this wish list request?  Unlike Oracle or SQL server database, we don't see any metadata information for tables in Kudu database (like primary key or indexed column), using an ODBC Impala connection Technical details : driver Cloudera ODBC Driver for Impala DSN v2.6.13.1015 64bitsdriver Cloudera ODBC Driver for Impala DSN v2.6.13.1015 64bits What is the improvement you would like to see?  Same behavior as the other types of database (Oracle, Sql server) : display key icon for primary key columns for example   Why is this idea important?  Displaying this information is crucial for advanced users querying our datalake database, in order to build coherent queries, espacially when joining tables, in order to avoid cartesian products and database server overload (timetout)   ... View more

What inspired this wish list request? I quite commonly have to use XPath to access JMP report layer and usually it requires using << Get XML   What is the improvement you would like to see?  I would like to have a feature which would allow me to easily view the XML of report. In best case it would be interactive, if I select display box from report, it would go to starting tag of XML, and in "worst case" it could be new script window/open to similar place as Show Properties.    Why is this idea important?  Makes it easier to see what is under JMP reports as XML. ... View more

  RE: Case Number 00060333/ 00055270   Brief Take Away: I'm requesting a drag-and-drop feature in Graph Builder to collapse multiple data points across a selected variable of interest [e.g. the "opposite" of overlay], with the goal of informing smarter SEM Error Bars determined by the N of the remaining number of data points displayed. This corrects multiple statistically misleading or incorrect visualizations and drastically increases efficiency that currently requires numerous work arounds that currently make exploring complex datasets challenging. This feature would make JMP incredibly more powerful.   What inspired this wish list request?  My name is Brian Sweis, MD, PhD. I am a psychiatrist and a behavioral neuroscientist working at Mount Sinai Hospital in New York City. I primarily work with mouse behavioral data generating very large and complex datasets. The tables I create in JMP to analyze my data have many different types of data including simple items like daily body weight measures or more complex task performance measures like trial reaction times tested on a decision-making task, where I may sometimes have 500 trials as separate rows per mouse x 40 mice x 60 days, easily getting to 1 million rows. I often am analyzing averages across mice, slicing my data in many different ways, including limiting analyses to a particular week of testing, or perhaps to only the first 10 trials of a given day or even multiple days, and so on. When I generate figures, exploring this vast dataset with Graph Builder fluidly is key. I often need to plot the Standard Error of the Mean on bar or line graphs, where in behavioral neuroscience, the desired error term for N samples is at the level of the number of organisms / mice. In datasets where I have numerous replicates or samples, particularly if it's a measure taken multiple times within a day or across days, and I want to plot group averages as described above, the SEM error bars are often incorrectly underestimated because N in graph builder is using the total number of samples in the display (e.g., sometimes N=5,000 samples if binned across days when really I need it to simply be N=40 mice).     What is the improvement you would like to see?    I stared a long thread conversation with JMP support team member via email. We ultimately decided together that the feature I would like would be extremely valuable if implemented correctly natively within Graph Builder and is not currently possibly without numerous work arounds involving having to generate many additional table summaries, columns of data, or scripts that quickly become untenable when the number of variables increases or if the ways the data are binned need to keep changing.   I will illustrate with a series of screenshots sample data, the problem, and a potential solution below.   First, here is example data for the bodyweight of 10 mice taken each day for 10 days, so N=100 samples total.       Next I will plot these body weights averaged across 10 mice, split by day when dropped in Group X on Graph Builder.       You can appreciate N=10 dots per day level. Here, average and SEM error bars are correct and as desired. This quickly becomes problematic below.     By simply removing day from Group X, all the data are polled together collapsing across days, but now we have all 100 samples displayed here. Each mouse shows here 10 times. Here, average is ok, but the SEM error bar uses N=100 as the error term, and not n=10 mice which almost always should be the case in behavioral neuroscience. It's a simple and common problem that would be incredibly useful if JMP builds in a native solution within graph builder. See below at how this gets more complex that can't be solved easily with current work arounds.         The obvious thing to do is use the Table Summary feature.       I often have to lean on this, but am usually left having to make hundreds of table summaries from my original dataset when I have more complex variables. When exploring my data, this gets out of hand very, very quickly. And it usually is in combination with a complex arrangement of having to apply data filters, making summaries of summaries, and so on and so forth.   One of JMP Support members suggested adding new columns where it calculates the average across the 10 days per mouse, and you can see below, this sort of works, but is quite cumbersome just for a single variable like body weight, and would not lend well to a table that has for example 100 other variables that are much more complex like behavioral task performance measures. That alone would add 8 columns x 100 variables and become unmanagable. Even extracting the formula for such columns and replacing the variable that the formula is using is not ideal and extremely manual / laborious.          This is especially true if I want to subdivide my data but multiple levels, especially if the levels are made up of asymmetric data points. And especially if I'm exploring my data in graph builder and am often changing these binning parameters on the fly. See below.      Here's another use case where I am also using local data filters to subset my data. This becomes evermore problematic to implement with scripts or new columns.        You can see that if I subset and include only days 7-10, on the left, when I split by day in Group X levels, of course it displays n=10 per bar just for the selected days as intended, but the pooled data on the right still has n=40.   Essentially, the optimal solution would be to implement a way to drag and drop a variable of interest in a new drop zone in graph builder. Below, I describe two possible solutions that I am imagining:   On the left annotated in red is a "mouse ID" centric method, either as a way to drop in mouse ID on graph builder and / or as an option to select in the error interval / error style option where the goal would be to reduce the 10 data points per mouse into a single data point per mouse natively all within graph builder here. Thus, we can tell Graph Builder at the end of the day what variable we want to determine the error term (e.g., via unique values, averaging across the remaining dots). On the right annotated in green is a "day" centric method - the variable I don't care about in this particular example and that I want to collapse across. Think of this as the "opposite of overlay" - instead of splitting by this variable, I want to collapse the data across this specific dimension (essentially the same as me intentionally leaving day out of the Table Summary grouping variable). I have a sense that the red-method makes more sense, but that the user may want to inform somehow which variable I'm explicitly trying to collapse across (e.g., day). Regardless, both would lead you to the idealized bar graph with the correct error term (n=10 total data points, 1 per mouse). When exploring data visually in graph builder, this could solve the more complex examples I describe above (e.g., with custom levels, data filters, etc).        Why is this idea important?    Having a rapid way to make smarter error bars adjust using my suggestion would making exploring complex data sets highly efficient if directly built into Graph Builder, natively. This corrects multiple misleading or statistically incorrect visualizations that often do not represent the underlying analyses many scientists and researchers perform when using individual organisms as the assumed error term. This avoids numerous work arounds that either quickly become cumbersome, untenable, or impossible to manage with larger or more complex datasets. This level of control in Graph Builder would be unprecedented and I think would be a unique and incredibly powerful tool.    ... View more

What inspired this wish list request?  2つの異なる母集団のデータの同等性を統計的に記述したい。頻度論的な統計学では同等性の検定の限界があるのでベイズ統計を使用したい。一般的なベイズ統計は複雑な積分を解く必要があるためMCMC法を搭載したアルゴニズムを搭載する必要があるが一部の手法は簡潔化されており複雑な積分を解く必要はない。下記に簡潔化したベイズT検定の例を示す Rouder’sMethod G¨onen’sMethod Hyper-PriorMethod   以下文献です [G¨onenetal.,2005]G¨onen,M.,Johnson,W.O.,Lu,Y.,andWestfall,P.H.(2005).Thebayesiantwosamplettest.TheAmericanStatistician,59(3):252–257. [Kruschke,2013]Kruschke,J.K.(2013).Bayesianestimationsupersedesthettest.JournalofExperimental Psychology:General,142(2):573. [Patil,1965]Patil,V.(1965).Approximationtothebehrens-fisherdistributions.Biometrika,52(1/2):267271. [Rouderetal.,2009]Rouder,J.N.,Speckman,P.L.,Sun,D.,Morey,R.D.,andIverson,G.(2009).Bayesian ttestsforacceptingandrejectingthenullhypothesis.Psychonomicbulletin&review,16(2):225–237. [WangandLiu,2016]Wang,M.andLiu,G.(2016).Asimpletwo-samplebayesiant-testforhypothesis testing.TheAmericanStatistician,70(2):195–201.   What is the improvement you would like to see?      Why is this idea important?      ... View more

I don't have a lot of faith in the 'plot residuals by normal quantiles' for doing model diagnostics. I have typically saved the residuals from the model to a new column, then used the Distribution platform to create a normal quantile plot there.   The fit of the solid red line does not seem to match up well with where the residuals fall using Fit Least Squares. The slope of the solid red line is not a good fit. I looked at a few sample data sets and saw issues with the default model scripts in these data sets:   2x3x4 factorial Bounce data Wine data Below is a plot of the 2x3x4 factorial Residual Normal Quantile Plot. I've been told that the slope of the solid red line is based on the RMSE of 2.02.     If I save the residuals to a new column and regenerate the Normal Quantile Plot from the Distribution plot, the line is based on the standard deviation of the Residual Y column of 1.03. It looks like a much more sensible fit from a model diagnostics perspective.   Is this something that can be updated in future versions of JMP? Note that when you use Fit Mixed, the results in the Residual Plots there match my desired result from manually saving the residuals to a new column. ... View more

☐ cool new feature ☐ could help many users! ☑ removes a „bug“ [~ lack of security] ☐ nice to have ☐ nobody needs it   What inspired this wish list request?  If the Database driver doesn't allow the user to store the password in the ODBC settings, it is difficult to store database passwords in a secure and simple way.   Very insecure: store the database password as plain text in the JSL and table (!!!) script   With Jmp17 ,there is the new option to Hide ODBC Connection String which leads to the effect that  Database table scripts (Update from Database, Modify, etc.) or .jmpquery files stop working    Another option: Encrypt the code for the database connection:  JMP-and-Oracle-Tips-and-Tricks-for-a-Happy-Marriage But then it is essential to use a Run Password. Otherwise, everybody with access to the code has access to the data base. And the run password just shifts the problem - and makes it more difficult to automate it.   The best approach: Store the Password at a save location and give Jmp access  to the password: Set and get credentials in the Credential Manager via PowerShell on Windows  Unfortunately, this approach is not compatible with Query Builder.   What is the improvement you would like to see?  Include a credential manager / connection to Windows credential manager in the Query Builder to automatically store database passwords in a secure, well-defined and simple way. e.g. along the idea of @ih  Set and get credentials in the Credential Manager via PowerShell on Windows    Please add a functionality to Create Database Connection() and New SQL Query( Connection()) such that they can use the stored passwords as well.   Why is this idea important?  A secure, well-defined and simple way to manage credentials - this is where standardization can help a lot. The best: directly include it into the existing Query Builder ecosystem.         more wishes submitted by   ... View more

What inspired this wish list request? Other user messaged me and told that they couldn't save a project and it did take some time to figure out why it couldn't be saved (most likely due me not using projects)     What is the improvement you would like to see? Improve the content management within JMP Projects: Provide easy way to move all files which aren't currently self-contained into the self-contained project Allow saving project as self-contained I think currently you have to first figure out which aren't currently self-contained and then save them all from Save As menu Allow user to see which file is self-contained and which is not Can you tell me which of these are self-contained and which are not?  Maybe even provide a preference where projects would make all content self-contained automatically   Why is this idea important? Improves one of the things JMP Projects are supposed to be used for and good at: content management.     ... View more

Hi.  I've been writing scripts to automate visualization of distributions, fitting distributions, and reporting cpk.  I just finished a long one, and really found a lot of disparities in how the names of some distributions must be used.  This required a significant amount of debugging.  My ask is not a small one, but I would love if JSL could make usage of distribution names more consistent in order to make writing scripts more efficient and less confusing.  I will provide examples below.   My script takes a list of distribution names {"Exponential", "Normal 2 Mixture", "Johnson SU", ...} and first generates a table of random data with one column for each distribution.  It then creates distribution plots, fits the data, extracts the fit parameters for each distribution.  It also uses the Process Capability platform to compute cpk.  I rely a lot on creating JSL Quotes then substitution of the desired distribution name to get the exact code I need.   Issues: When I want to generate matrices of random data initially, I need to use "Exp" instead of "Exponential" and "Normal Mixture" instead of "Normal 2 Mixture".  J(100, 1, RandomExp(...)); J(100, 1, RandomNormalMixture(...)); J(100, 1, RandomJohnsonSB(...)); On the other hand, to fit a distribution to my data, I need to use "Exponential" instead of "Exp", "Normal 2 Mixture" instead of "Normal Mixture" and "Johnson" instead of "Johnson  SB": dt << Distribution( Continuous Distribution( Column(:colname1), Fit Exponential, Fit Normal 2 Mixture, Fit Johnson ) )  And again another naming scheme for the Normal 2 Mixture in the Process Capability platform: pc = dt << Process Capability( Process Variables( :colname1 & Dist(Exponential), :colname2 & Dist(Mixture of 2 Normals), :colname3 & Dist(Johnson), ) ) I had initially hoped I could just provide my script a list of distribution names to try, then have a for loop cycle through the list, adding the specific name to a JSL Quote, and then executing.  Instead I found out the hard way that I needed to create several if clauses to fix the distribution names for each different operation prior to substitution.   It would be great if all of this could be simplified by just using one set of distribution names for all functions and platforms.  Thanks.   [Using JMP Pro 17.2.0] ... View more

What inspired this wish list request?  Currently, when I trying to chose some samples from a plot of graph, it can only select by drawing a rectangular.  However, a lot of times, I need to select samples by a circular or oval shape, it's very hard to cherry pick each sample by using current method.    What is the improvement you would like to see?  Add a shape chose for select arrow, so we can chose if we want to select samples by drawing a circle or rectangular.    Why is this idea important?  Many company working on wafer like products, it's naturally a circle shape.  In this case, a lot of times, there are circular or oval shape defects, uniformity issue that need to be selected.  The user group is huge, since most semiconductor or other circular products company or manufacture need such tool to help.  It will be greatly saving time to have such selection shape.     ... View more

In fit parametric survival platform only 2 options of equation can be saved: save probability formula save quantile formula In majority of cases users of fit parametric survival platform are interested in equations of upper or lower confidence intervals for probability or quantile case.   I propose to add following save formula options to fit parametric survival platform: save lower/upper confidence intervals probability formula save lower/upper confidence intervals quantile formula   P.S. The names in fit parametric survival platform are inconsistent: time in profilers and Estimates of Survival is called time, while in Estimates of Quantile it called quantile probability failure in Estimates of Survival and P in Estimates of Quantile In fit parametric survival platform I propose to change: name of quantile to time  in Estimates of Quantile name of parameter P to probability failure         ... View more

What inspired this wish list request? We are using this as an introduction to JMP class. It would help if people can get completion certificate and send it to a contact when they complete the course.     What is the improvement you would like to see? It would help if people can get completion certificate and send it to a contact when they complete the course.     Why is this idea important? It would help if people can get completion certificate and send it to a contact when they complete the course.     ... View more

What inspired this wish list request?  I wanted to compute the uncertainty of prediction fastly and easily for predictive model without using bagging. Indeed, bagging is quite slow to compute with the profiler.   What is the improvement you would like to see?    I would like to save the all the trees formula when I use the Bootstrap Forest Plateform. Or at least the standard deviation of the distribution.     Why is this idea important?    This is very important for me to be able to compute acquisition function when I do Surrogate Model Optimization (Bayesian Optimization). I tried to use bagging with the partition plateform, but the result is not satisfactory. The gaussien process plateform do not deal with categorical variable.     ... View more

What inspired this wish list request? When using this platform, if one or more variables are continuous, the analysis gives you a message to change the variable.     What is the improvement you would like to see? Since the platform knows the problem, it should offer me a way to temporarily or permanently make the change to the variables(s) without having to stop thew analysis. Make the change, then come back.     Why is this idea important? Will make the analysis more efficient.     ... View more

What inspired this wish list request? I was at the Greater Philadelphia Area Pharma JMP Users Group Meeting today and learned about the existence of the wish list from the JMP staff. I had requested this to a JMP representative at a different event a few years ago who told me to call JMP Tech support to make the request, but I never did.   I have "Box Plots" and "Plot Actual by Quantile" on by default due to my Platform Preferences for Fit Y by X with a continuous Y and nominal X. (This side-by-side plot is my favorite graph in JMP.) The red default is confusing because it does not match the legend color.   So, I usually change it by right clicking, selecting Customize...Customize Graph... then repeating Box Plot, Box Plot Line Color, and changing the color for each level of X.   Default red boxes misaligned with the legend   Color scheme aligned with the legend that comes with the Q-Q plot.     To make things worse there is a bug in JMP that I have not reported before, which I will report separately to JMP Technical Support. After I change all the boxes to black and save the graph to a script Window or a script in the data table my color customization does not stick. So, I have to redo my customization every time I make the graph. I am using v17.0 now, but I think this has happened as long as I can remember (>15 years). Today is the first day I figured out that changing the color to anything but black sticks in the script because I generally use black. (I know I could add the command lines to the script manually and resave it.)   All black boxes (not saved properly in a script):   What is the improvement you would like to see?  In Fit Y by X...Oneway, Box Plot, the box color should have an option to be other than red in the Platform Preferences. One choice could be black. Another could be the same color scheme used by "Plot Actual by Quantile".   Why is this idea important? Red is confusing for side-by-side box plots when the legend for the normal quantile plot has red to indicated the group with the first level of X. Red is confusing when the dot plots under the box plots have different colored symbols, and one of them is red. Black would work because other elements of the graph are in black.  The same color scheme used by "Plot Actual by Quantile" would work, especially when the Q-Q plot is part of the same graph to the right of the box plots and using the same Y-axis as the box plots.   ... View more

☐ cool new feature ☐ could help many users! ☑ removes a „bug“ ☐ nice to have ☐ nobody needs it What inspired this wish list request?  from @ https://community.jmp.com/t5/Discussions/How-to-know-if-a-table-is-a-linked-subset/m-p/655280/highlight/true#M84437  If I try to concatenate tables with the option Append to first table, the script stops with the message:   Cannot append to table: Big Class. Possible reasons for this include: the table is locked, the table is in a platform with a by group, or there is a linked subset of the table. in access or evaluation of 'Concatenate' , Concatenate(/*###*/Data Table( "Subset of Big Class" ), Append to first table)     The error message was very helpful to identify the issue. But how to make the script robust against linked subsets?   What is the improvement you would like to see?  An addition function: dt << get list of linked subsets() will help. Then I can close the subsets via JSL.     Why is this idea important?  code won't crash if there are linked subsets. Alternatives: Jmp updates the linked data tables Jmp cuts the link automatically concatenate get a new option: Jmp_is_allowed_to_close_linked_data_tables_if_this_is_needed_to_run_this_concatenate(1)        more wishes submitted by   ... View more

☐ cool new feature ☑ could help many users! ☐ removes a „bug“ ☐ nice to have ☐ nobody needs it   What inspired this wish list request?  Every day I generate a dozen of summary tables, just with a single column as group variable and without any aggregation. ... just to get Jmp calculate the number of unique values in that column. Definitely much faster than writing a JSL code with Summarize :) But still: several mouse clicks.   And then I generate other summary tables just to get the mean, max, min ... value of a column.   I guess that other users use the Summary Platform for similar means.   What is the improvement you would like to see?  Add a functionality to show a short a column statistics in the status bar when the user moves the mouse over the column header histogram or the name of a column. In the preferences please proved a list with aggregation options - the more, the better.     Why is this idea important?  Huge speed improvement.   I "fear" that this functionality is already available (so obvious)  - and that I wasted my time by creating this post. where?         more wishes submitted by   ... View more