Power up Your Analysis in JMP Live 19 (2025-US-30MP-2528)

Hi, everybody. My name is Paul Spychala. I'm here with Praveena Panineerkandy, and we're going to be talking about how to power up your analysis in JMP Live.

Moving on. Why are we here today? We're here today. We're really excited to showcase, introduce some of the awesome capabilities and features that we've added to JMP Live.

It's really going to change what you can do in JMP Live in a way that you know has never been possible before. This is more on the analytical side to showcase more ways to explore your data. To talk more about that, I'm going to toss it back to Praveena.

Thank you, Paul. Let's briefly discuss and explore the new capabilities introduced in Version 19, especially in comparison to the earlier versions that preceded it. Prior to 19, users had encountered restrictions when it came to accessing platform menu options after publishing their work.

The only exception to this restriction was found in 18, in the Distribution platform. It was during this time that we began exploring the process of adding support for these features, which led to the development of a limited selection of menu options, which served as a stepping stone towards a more comprehensive functionality that we'll see in 19.

In 19, we are excited to announce that we now provide support to a significantly expanded array of platform menu options across three main platforms: Distribution, Bivariate, and One Way platform. In addition to this, we have also made improvements in enhancing the user experience by introducing the popular Fit to Window option found on JMP into the JMP Live environment. This feature aims to optimize the viewing experience and ensure that the content is displayed in the most user-friendly manner possible.

Before we begin the demo, let's just quickly go over the support for menu options that are available in these three platforms. In Distribution, we have support for Nominal and Continuous Distribution. We support some or all the submenu options that are available. In Bivariate, we support some of the platform menu for Bivariate platform, along with the child menu for the Fit Line. In One Way, we support almost all of the menu options along with the child menu of the Means Comparison menu.

Now let's see these features in action. First, we'll start with the Distribution platform. Here we have a report that is generated of the candy bars data set that is found in the sample data folder in JMP.

We have published a report that has one Continuous Distribution that is for the Calories column and one Nominal Distribution that is from the Brand column. Now, let's say after publishing, the user wants to increase or decrease the bin width, or see the count or percent associated with each of these bins, they can navigate to the menu options found next to Calories.

From there, navigate to Histogram options and do Set Bin Width. Here, the users can now enter the new bin width that they want and click on Okay. As we can see, the report gets updated with the new bin width in the Calories report, and if they want to see the percent or count associated with it, they can always turn it on. Users can now have the ability to add additional plots or graphs to the report that they have already published.

Let's say we want the Stem and Leaf plot, which are a great way to see the raw data while visualizing the distribution shape, we can just navigate to the menu and turn on the Stem and Leaf plot that is available. As you can see, they have a stem that is the leading digit, and the leaf that is the trailing digit for the each calorie value. This allows us to see the actual data points while also understanding the frequency distribution. Now we want to see, let's say-

Hey, Praveena. I'm so sorry. I think that you are trying to share an HTML page, you're trying to share JMP Live right now. You might only be sharing your PowerPoint presentation. You're only sharing that one screen. Sorry, I don't mean to interrupt, but obviously, yeah, you want to get that on the recording, right? Maybe just start back from… When you share in Zoom, it'll give you the option to share either an application window or you can share your entire screen.

That's what I would recommend is sharing that entire screen. That way we'll be able to view… There we go. Now I can see the JMP Live. No worries. But I will make note of when you went to the demo. Just start from scratch from the demo. Sorry, it took me a minute to pop in there. I was like, hold on. I will go back on mute, so restart whenever you feel like it.

Let's see this exciting new features in action. First, we have the Distribution platform. Here I have a report that is generated from the candy bars data set that we find in the sample data folder of JMP. We have published a report that has one Continuous Distribution based off the Calories column and one Nominal Distribution based off the Brand column. Now, let's say after publishing, the user wants to increase or decrease the bin width.

They can navigate to the menu and go to the Histogram options and select Set Bin Width and enter the new desired bin width, and click on Okay. As you can see, the report gets updated with the new bin width for the Calories distribution. The users, if they would like to see the percentage of counts associated with each of the bin, they can enable the show counts or percent within the menu option.

Now the report. We also have the ability to add additional plots like let's say the Stem and Leaf plot, which is a great way to visualize the raw data. Here we can see that JMP has now added the Stem and Leaf plot to the existing report showing the stem, the leading digit, and leaf, the trailing digit, for each calorie value. This allows us to see the actual data points while also understanding the frequency distribution.

If we have to see the proportion of data that fall at or below a certain value, we can enable the CDF plot. JMP will add a new graph that shows the cumulative percentage on the y-axis and the calories on the x-axis. We can see, for example, what percentage of candy bars have a calorie of 200 or less. For our distributions, like having a continuous variable, JMP can fit various statistical distributions to our data to see which one it closely resembles.

From the menu, we select let's say, Normal Fit. We can see JMP has added the Normal Fit Distribution report to the existing report that shows the standard deviation and mean, and it also has a distribution curve added to your existing report. We can also enable profilers associated with this distribution. If we would like to see a different distribution, we can always enable another distribution, let's say lognormal. We can compare these distributions within our report.

Now let's say for Nominal Distribution, we want a visualization that gives a quick and a single column representation for the proportion, making it easy to see which brand makes the largest and smallest percentage of the data set. We can enable the Mosaic Plot. These are some of the main Distribution menu options that are available on JMP Live.

Moving on. The next one is the Bivariate report. Here we have a report that is generated of the horror movie data set from Kaggle. We are trying to see if runtime has any effect on the voting average. Now, in this report, I can turn on the histogram borders. Doing so will result in having two histograms added, one on the top that reflects the histogram associated with the runtime, and one to the right that is associated with the voting average.

If you look at the histogram for runtime, we can see that the peak is around 80–100 minutes. Then it tapers towards the shorter and longer films. This helps us to understand the range and commonality of movie lengths.

Moving on. Looking at the voting average, we can see that horror movies are not mostly critically panned. Large number of them have a low rating, and only very few horror movies achieve a higher rating.

We can also enable summary statistics if you are interested in analyzing any other statistical information associated with these variables. We can also add Fit Line. Fit Line helps us quantify any linear regression. Here we can see that our JMP has added a regression line in the scatter plot. Below, we can see a report that has all the irrelevant statistical detail associated with the Fit Line.

Each of these Fit Line has their own menu options. Let's say we turn on the Confid Curves, so we can see that in here JMP has added two dashed lines. It can tell that for any given runtime the range within which can be like 95% confident that the true mean of vote average of all the movies are within the runtime.

If we notice, at the mean, that is where it narrows, and it widens at the extremes, reflecting that it is less certain at the extremes. Next to the Fit Line, if needed, we can remove the Fit Line, or if needed, we can add multiple Fit Lines. These are the main menu options that are associated with the Bivariate menu. Next, I hand over to Paul to demo the menu options with the One Way platform.

Thank you, Praveena. Here we are with another Halloween-themed data set. This is something we also found on Kaggle. This is a study where people were voting on which is going to be the best candy or their favorite candy. We're giving a choice of two candies. Then it would basically say, which one would you rather have?

They did a whole bunch of comparisons. Hundreds of thousands of people voted. The one that they picked would be considered the winner. This is showing the win percentage. Here for this data set, we have Reese's Peanut Butter Cup. That's a top candy here, for chocolate candy. It basically won 84% of time. If you had Reese's Peanut Butter Cup with some other candy, no matter which one on average, you won 84% of the time.

For candy that does not have any chocolate, the top one is Starburst, and it basically won two-thirds of the time. We're trying to find out, hey, does chocolate seem to have any effect on whether it is a really popular candidate? If you are going to make your own type of candy, if you are in a candy producing business, maybe this analysis will help you figure out, hey, should I make chocolate-based candy or candy that has sort of chocolate in it?

But let's go jump in with the new analytical capabilities we're adding in JMP Live. This is also down to the red triangle menus in JMP. In JMP Live, it's got these triple dot menu items, all of it's under here. We've added all these new statistical test methods, visualizations, all kinds of things.

We can just start out with the first one. This is the quantiles example. Sorry, statistics. You can easily add that to your report if you want to add the Means, Anova, or maybe the Pooled t Test. You can just add it… If your report already had it, you can also take things out. This is just maybe too much information. I don't want this anymore. You can also take them out of the report. We have Analysis of Means, maybe something you want to add to showcase the difference of the means.

You can also compare means. You can do each pair. You can do all pairs. It also adds submenus in here, so you have the confidence quantile. If you don't want some of these, you can turn them off. You can disable, like, hey, we don't want each pair. Let's get rid of that just to have all pairs. All of these options are available. I'm not going to showcase everything. But you can add non-parametric tests.

They get added to the bottom. Just like in many other plots, we can add a CDF plot cumulative distribution function, and that gets added to the bottom. It looks just like the CDF plots of other platforms. We can add the Normal Quantile Plot. Just like with a lot of customizations in JMP, a lot of them now are available in JMP Live.

Here I added histogram. Shows the distribution of candy that has chocolate and candy that does not have chocolate. Maybe if you want to make the perfect candy, some of these analysis in One Way will help you do that. We're going to move to a different addition. That is the Fit to Window.

JMP Live is unique. As in, you don't really have as bit of control of who might be viewing it, just because it's meant to be shared with other people. Somebody might be looking at this report on their mobile phone, and it's a very small screen. So if you size the report to be a specific size for like a laptop screen, and then somebody views it on an iPhone or a mobile phone, they might have… The window might be smaller and might be too small. But now, these reports should automatically resize to your screen.

If you have this on a TV, like in a conference room or something like that, you can stretch out, have a multi-monitor setup, and I'm stretching out across multiple screens, and it works fine. Similarly, all this is available in this triple dot menu. You can just turn it off like, hey, I want this to be this size, and no matter what my screen size is, it's going to stay that size. Otherwise, if the proportionality of the graph is important, and no matter what the screen is, it's going to maybe become smaller or bigger, but the proportions, the aspect ratio of the graph is going to stay the same.

Let's go turn it back to on. Anytime you use any of these size settings, the graph is going to shrink and grow to match that size. That's basically it for this one. I think it's really going to help people size the reports or just leave that decision of, "Hey, how to make it fit nicely on my screen," out of your thoughts because it's going to be handled automatically. This will be only handled in Graph Builder and Control Chart Builder.

Platforms are really long and, I don't know, meant to be consumed like a report where you're scrolling through pages and pages. We're not going to try to fit those because they're meant to be scrolled through. They're meant to be read across maybe multiple pages. We're going to close this up. Here are some of the resources that are available in our presentations. You should be able to download them off the presentation website. I want to just thank you for watching our presentation about how you're going to power up your analysis in JMP Live. I hope you guys are excited about some of these features as we are.