JMP Wish List

Suggestion would be to allow an automatic solve for the difference value that gives equivalence for a given alpha. In the example example-of-an-equivalence-test there would be an option to solve for the difference value that predicts equivalence for a given alpha value (it addition to the current option for supplying a specific difference).   Many times this difference value is specified inconsistently among a given department or functional group (often with many differing opinions).  So, instead of running multiple differences to show the p-values for equivalence it would be very convenient to solve for the difference that would predict equivalence given an alpha value. ... View more

In data tables with some more columns it can be useful to preselect rows to tidy up and document what you're doing. Then one would know later which cols have been selected for e.g. predicting.   But I'm not able to use this feature, because starting a platform with existing preselected rows (x and y) usually lets the platform run directly. It is then not possible, to add e.g. some by cols etc. So my way currently is, not to use preselected rows, and Group cols accordingly (to remember which are my x's).   I think it would be better to stop platform at launch window even in case of preselected rows. Or at least configure in Preference which behaviour is preferred. But I would prefer the first Option.   There's a thread regarding this behaviour, and the presented solution (relaunch) is not helpful for platforms that take some more time to run ... https://community.jmp.com/t5/Discussions/Preselected-Role-and-automatic-Analysis-How-to-stop/m-p/11928   ... View more

Add options to change how bars are displayed on histograms within graph builder, including stacked, side by side, nested, etc.   Users would find it more intuitive to build this chart by starting with a histogram rather than needing to use a bar chart with a count statistic:   but the closest you get are overlaid bars:   Code: Names default to here(1); dt = Open( "$Sample_Data/iris.jmp" ); dt << New Column("Sepal length Binned", Numeric, "Continuous", Format("Fixed Dec", 12, 1), Formula(Floor(:Sepal length / 0.2) * 0.2)); dt << New Column("Is Target", Character, "Nominal", Set Property("Value Colors", {"Is Target" = 59, "The Rest" = 32}), Formula(If(:Species == "setosa", "Is Target", "The Rest"))); //Desired Chart dt << Graph Builder( Size( 500, 300 ), Show Control Panel( 0 ), Variables( X( :Sepal length ), Y( :Sepal length ), Overlay( :Is Target ), Color( :Is Target ) ), Elements( Bar( X, Y, Legend( 11 ), Bar Style( "Stacked" ), Summary Statistic( "N" ) ) ), SendToReport( Dispatch( {}, "Sepal length", ScaleBox, {Min( 4.1659880239521 ), Max( 8.0340119760479 ), Inc( 0.2 ), Minor Ticks( 0 )} ) ) ); //Histogram option - no way to stack bars, leads to confusion when 'Is Target' //and 'The rest' have similar counts dt << Graph Builder( Size( 500, 300 ), Show Control Panel( 0 ), Variables( X( :Sepal length ), Y( :Sepal length ), Overlay( :Is Target ), Color( :Is Target ) ), Elements( Histogram( X, Y, Legend( 12 ) ) ), SendToReport( Dispatch( {}, "Sepal length", ScaleBox, {Min( 4.1659880239521 ), Max( 8.0340119760479 ), Inc( 0.2 ), Minor Ticks( 0 )} ) ) ); ... View more

Most people who manipulate 3D scatterplots do not want to rotate about the Z axis, and they end up struggling to keep it 'right side up' after a few rotations.  Make the mouse movement work like the 'Move and Rotate' behavior in the animated gif below.            Here is an add-in that provide the functionality in the animation above:      https://community.jmp.com/t5/JMP-Add-Ins/3D-Plot-Tools/ta-p/48699   Proposed Provide an option to lock the z axis, enable it by default.  When enabled, horizontal movement of the mouse rotates the chart about the y axis, no matter where the user initially clicked on the chart. Vertical movement rotates the chart about the x axis. ... View more

For those who use "matching" techniques, e.g., propensity score, mirror histogram is a common way of visually showing how similar or different a distribution of a continuous variable is between classes (categorical variables). For publication, we believe, it must be MIRRORED. I draw mirror histograms using my custom made script and want an easier way. One way of making it possible to draw a mirror histogram as a standard feature may be to add a category-specific "flip axis" option in Graph Builder platform. ... View more

When I do a contingency analysis, the numbers in the table are too big for the column and they end up overlapping. There is no way to adjust the size of the columns or the text formatting so that this doesn't happen.     ... View more

Hi all,     So I'm using the graph builders to display pictures. You know, the "expression" data type. The DIY solution that has been implemented to embed pictures into a JMP table.   I want to see these pictures clearly when I'm looking at them on graphs, so I use a marker size of 100 (I know, I'm just looking for trouble). Now, I want to see the labels for these pictures when I hover over them. But then nothing happens! Then I tried to find the exact center of the pictures and it works, the label displays! That means, to see the label, you have to find the small area that corresponds to a default size. From what I can see, this area is a circle of about a marker size 7, regardless of your actual marker size. Even if you use size 0, the label will display in that area.   So my request is as follows : Make the label to display on the whole size of the marker, whether it be a normal marker like a dot or something more fancy like a picture. ==> This should also take into account the shape of the marker. By default this is a circle but this should change with the shape of your marker, like a rectangle for pictures.                   ... View more

I am making Pareto/ mosaic charts and I would like to add more than one cell label, e.g. having percentage AND number of observations. So in the graph builder, it would be a series of tick boxes instead of a drop down menu to choose this.   I have tried doing this by adding a "cartridge plot" (the first one on the right of pareto plot),  as well as the Pareto chart in the graph builder but it just doesn't work well. The scale is resized in a random fashion and the graph becomes unreadable.     So far my solution is to make two charts, one with the percentages and one with the number of observations and put them both on the same slide in powerpoint... ... View more

I make a lot of column/bar graphs to represent data. Then I run an analysis which evnetually leads to a Tukey's Mean seperation. I would love the ability to label the graphs with the output of the Mean seperation in a quick and easy manner! ... View more

One change I don't like very much in JMP15 is the use of histogram based control for continuous data variables.  There are two aspects that I don't really like: - the vertical reference lines for the upper and lower limit are difficult to select.  They present initial at the range of the histogram display, and mouse control is tricky to select them.  I find that I too often select the border of the display element and when I attempt to slide the control I actually resize the histogram display. - It is not directly obvious that you can select the numeric value below the histogram for the upper and lower limit to change it.  The number display has no highlighting or border to indicate it is selectable or changeable. - I miss the slider bar.  Please bring it back as an option, with our without the histogram display   Another thing I would like to see is to be able to easiy treat a continuous variable as categorical in the data filter.  For instance, in a DOE, the DOE factors are often at discrete levels on purpose, and if I want to filter the results of the table based on the DOE factors, I need to change the column modeling type to nominal or ordinal so that I can select discrete values.  Then I would need to later switch the column modeling type if I want to do modeling on the DOE results.   Often what I want to do is just select on level of one factor in the DOE, and doing that with the continuous data filter is clumsy.  Maybe a "discrete continous" filter could be added that allows choice of the discrete continous value(s) without having to change the modeling type. ... View more

I would like the ability to get alias effects matrix using JSL in Custom Design.   there is a <<Get X Matrix  (X) that returns the design matrix for the specified model and there is a <<Get Alias Matrix message that gets the aliasing correlations (for model x alias interactions, X`*A) but there is not command to get the alias effects matrix for the design (A) In Custom Design, Evaluate Design, and Compare Designs, I would like to be able to use JSL get the full correlation matrix that is used to construct the color map on correlations.  So a <<Get Correlations Matrix message would be nice.to add.   For Evaluate Design, I would like to be able to see a list of the pairwise correlations of the model x model, model x alias, and alias x alias correlations.  Similar to the pairwise correlations table that can be displayed in Multivariate.  The reason so so that I can sort and see the pairs of effects with the highest confounding.     For Compare Designs, I would like to be able to see the difference in the effects and aliases correlations matrices.  So if I compare 2 designs, and each has a color map on correlations (C1 and C2), I want to be able to see (C1-C2) as a "delta color map".   ... View more

  Hello, Would it be possible to create a simple implementation in Graph Builder or the Distribution Platform to easily plot two histograms and calucate the proportion of area shared by them under a particular distributional assumption? (eg. Normal)   https://community.jmp.com/t5/JMPer-Cable/How-to-overlay-histograms-in-JMP/bc-p/229894/highlight/true#M386   Minitab has a decent implementation of this: I copy-paste a graphic of their output here for your reference. Also the link to the above post shares the specifics on the ask in more detail per my discussions with @XanGregg and @Steven.    cheers, @PatrickGiuliano  ... View more

Currently, under Win10 & JMP 14.3 Pro, it appears that the only way to reorder independent variable terms within the Construct Model Effects box in the Fit Model dialog is to remove them, then re-add them in the desired order. I would like to suggest allowing their order to be shifted up/down simply by clicking and dragging within the box. This would be especially helpful for fitting models using Type I SS (where predictor effects depend on the order in which they are entered into the model). ... View more

Currently (in JMP 14.3), Value Ordering is not respected in Graph Builder - Heatmap. ... View more

I make a lot of charts for presentation in Apple Keynote and MS Powerpoint. I use Graph Builder quite frequently. Very often, I have to go back and modifty a chart some time after I created it. When I do this, I quite often have no idea what table the chart was saved with.   It would be wonderful if Graph Builder (and perhaps other JMP platforms) would have an option to add text to a chart with the name of the source table. Something small and not too visible in a lower corner perhaps. This would vastly simplify tracking down charts when I need them again.   Thanks for all the great work you do. ... View more

I strongly request JMP please develop a more completed map of China for Chinese users in the upcoming version, including Hongkong, Macau, Taiwan, islands of South China Sea and nine-dotted line so that I can publish my analysis result to some platforms like video website, BBS, and even knowledge database website to publish my paper. Ref: https://community.jmp.com/t5/Discussions/Issue-about-JMP-s-built-in-map-of-China/m-p/186279#M40472 ... View more

Can you add in preference settings for the OpenGL View Settings (accessible currently via right click in Scatterplot3D).  I find that the settings for Marker Size are way to small for me to effectively visualize data, especially small data sets and DOE factors, so I would like to be able to set the view settings to have a marker size that is larger.     Also, it might be nice to the marker size setting for the graph be adaptive to the size of the data set and the distance between points.  Seems like there is some optimal visualization setting, where the size of the point should be related to the distance between the points.  The more/less dense the data is over the range of the data, the smaller/larger the data point marker size should be.   ... View more

The distribution calculator is a great tool for teaching and to look up certain quantiles. Other tools (e.g. 'Confidence Interval for One Proportion' or 'Sample Size for Confidence Intervals') are also a great help, but they're hidden away under the Help -> Sample Data -> Teaching Scripts / Calculators menus.   I would love it if they were more easily accessible.   There are add-ins available with these contents, but they don't always work as reliable as when going through Help -> Sample Data -> ...     ... View more

I would like the option of choosing different statistics besides the mean (i.e. standard deviation, quantiles, range etc.) when using a coloring variable in the graph builder heatmap visualization. In addition, when the heatmap cells have a count = 1, it would be nice to have the heatmap cells colored by the coloring variable column attributes: value color, value label and value order when defined. This would be especially useful when the coloring variable is nominal, like in the case of using bins to color die in semiconductor wafer maps. If there is a way to do this via JSL I'd be interested to learn more about it.    Thanks,   Francis ... View more