Thursday, February 4. 20109-5 vs. 5-9
You’d have to be living in the most desolate areas of the globe not to feel the onslaught of information clamoring for our attention. In the midst of it all, there are pieces that inform and enlighten, but most information is just noise. Those of us who spend time on our computers from 5 p.m. to 9 p.m. are able to take advantage of some very savvy, user-friendly, semi-efficient ways to cut through the clutter.
For example, I love Google Finance’s interactive stock charts and Bestiario’s Videosphere project, showing the semantic relationship between videos. But from 9 a.m. to 5 p.m., many of us are saddled with applications that were developed for much smaller volumes of data and that take much longer to respond to our information needs. Most information research states that data is doubling every 18 months. Some say even faster. Regardless of the actual timeframe, data is growing faster than it can be consumed and digested. Companies and individuals are constantly being challenged to make decisions more quickly with less time to analyze the data that is required as supporting evidence for their conclusions and proposed courses of action. What makes 5-9 applications so much more appealing? Much of it is their ability to present information visually and use graphs that the viewer can interact with. But the 9-5 world is catching up. It’s moved from data-heavy reports, to static graphs, to dashboard widgets, to today’s interactive graphs (many of them copied from the Internet). It’s been well documented that one of the best ways to understand data, especially lots of it, is through good data visualization. Since most of us at JMP spend lots of time on our computers (much to the displeasure of our significant others), we are constantly on the lookout for better ways to present information to our users in an interactive, intuitive, visual way. We know we can learn a lot from 5-9 applications. What are your favorite 5-9 applications and why? Monday, February 1. 2010Keyboard Tips: CTRL/Click , ALT/Click, CTRL/ALT/Click
In his January 28 Webcast on Advanced Tips and Tricks for Maximizing Your Efficiency, Sam Gardner demonstrated three keyboard shortcuts that are useful for applying commands or options to graphical reports.
1. CTRL/Click applies one red triangle hot spot command or option to all applicable graphical reports in an analysis window. For example, in a display of multiple Distributions, you may want to see the Normal Quantile Plot for all continuous variables. Simply hold down the Control key while you click the red hot spot and select Normal Quantile Plot. JMP is smart enough to identify the Distributions for the continuous variables and display the Normal Quantile Plot for each. ![]() 2. ALT/Click applies multiple red triangle hot spot commands or options to one graphical report. For example, you may want to Fit Line, display Histogram Borders and Fit Mean for a scatterplot. Simply hold down the Alt key while you click the hot spot and then JMP displays a checklist of options. Check the options you want, click OK, and JMP applies all the selected options to the graphical report . ![]() 3. Want to combine both shortcuts? CTRL/ALT/Click applies multiple commands or options to all applicable graphical reports in an analysis window. Want more tips? A recorded demo that includes other advanced tips and tricks for using JMP is available for download. . Want to join a live Webcast? Webcasts offering demos and helpful tips for JMP users are held at least once per week into May. Register for as many as you like. Wednesday, January 27. 2010The Pictorial Superiority Effect with JMP
JMP benefits heavily from the pictorial superiority effect. To save you a moment of searching in Wikipedia, here's what that is: "According to the picture superiority effect, concepts are much more likely to be remembered experientially if they are presented as pictures rather than as words." I was reminded of this effect when a friend sent me a link to Alex Lundy's talk on "Chart Wars: The Political Power of Data Visualization." It is a rather humorous (although please be warned that the video contains some expletives) and educational talk espousing the benefits of tools such as JMP.
In the video, Lundy displays the names of a few of the experts in the field of data visualization. We pay a lot of attention to these data viz gurus at JMP; specifically, you may have noticed that we frequently invite Stephen Few to speak at JMP events. I would add Ben Schneiderman to that list, and there is certainly a more comprehensive list to compile. Lundy also shows a quote from H.G. Wells -- "statistical thinking will one day be as necessary for efficient citizenship as the ability to read and write" -- and adds the word "visual." Add the word "dynamic" to this statement, and now you are describing a state-of-the-art statistical tool called JMP. To demonstrate this, let's use his example of published White House staff salaries. You can get the data set yourself. I was able to generate the same graphic Lundy shows and distribute it in a few minutes with JMP. Using a combination of the Distribution platform with the Data Filter and Tabulate, I quickly gleaned more information about the sum and distributions of salaries of different subsets of the White House staff. ![]() For example, the combined salaries of staff with the status "Detailee" is $3,966,222.00, and they have significantly higher mean salaries ($123,949.19)... ![]() ...than the rest of the White House staff ($77,320.55), who have a combined salary of $35,180,848.00. ![]() You can easily get this information in JMP, and it is not trivial to do this with other software. This is why I believe JMP is the best statistically visual software product on the market. Tuesday, January 26. 2010Looking Back and Ahead to JMP Academic Webcasts
It’s that time of year for academics, the start of a new semester. The JMP academics team is busy preparing for the academic Webcasts we deliver at the beginning of each semester. We thought it would be interesting to share with you our results from our fall Webcasts. We delivered four sessions called “JMP Basics for Professors and Students,” and 148 people attended with a 98% satisfaction rating.
The 148 people were made up of 28% each graduate students, research faculty and teaching faculty. If we combine the categories for students and teachers, we can see that 90% of our attendees were students, teaching faculty and research faculty. Somewhat surprising to us was that the No. 1 discipline among attendees was life, health or physical sciences, followed by general and social or behavioral sciences. These three groups made up 77% of our audience. In academics, we have traditionally had many users in the business and economics group as well as the engineering or computer science group, so we were happy to see so much interest from people in the first three groups as well. Other fun facts about our attendees include that 28% were Mac users and 72% were Windows users. About 70% said they had little JMP experience. And we had attendees from 32 states. Starting on January 28, we will be hosting the same Webcast as in the fall, “JMP Basics for Professors and Students,” as well as a new Webcast called “Teaching Statistical Concepts with JMP.” “JMP Basics for Professors and Students” will be offered on Jan. 28 and Feb. 9, 12, 15, 19 and 24 at a variety of times. This Webcast covers an overview of basics for teaching or using JMP for elementary and intermediate statistics courses. “Teaching Statistical Concepts with JMP” will be offered on Feb. 10 and 23 at 1:00 p.m. EST. This Webcast covers the resources available to help teach statistics using JMP. You can register to attend one or both of these. Hope you’ll join us and spread the word!
Posted by Melodie Rush
in Academic, JMP - General, Statistics, Training
at
10:33
| Comment (1)
| Trackbacks (0)
Friday, January 22. 2010A Tidbit About Tabulate
During yesterday’s Mastering JMP live Webcast on Shaping Your Data, one of the JMP capabilities presenter Jeff Hawes demonstrated was Tabulate (Tables>Tabulate). Here is a tidbit that folks found interesting.
Jeff: Tabulate is one of the data management tools in JMP. It is akin to a Pivot Table in Excel, but more interactive. By dragging and dropping into zones on a template, you create a multidimensional table. Question: What is the advantage of using Tabulate instead of Summary (Tables>Summary)? Jeff: There are two main differences. First, Summary only summarizes for rows - and the table is made up of records in rows with single-header columns. With Tabulate, you can group your rows and columns into subcategories and nest factors within multiple categories. For example, you can build subcategories like vehicle size for each car brand. Second, Summary provides a data table, whereas Tabulate is a an interactive report that responds to dynamic, animated data filtering. In this example I apply the data filter to highway mileage. ![]() Want to learn more? • View a recorded demo on shaping your data. • Register for any of our weekly live Webcasts in the Mastering JMP series. • If you have JMP and want to play with Tabulate using the same data Jeff used, get Cars 1993.jmp from your JMP sample directory. Thursday, January 21. 2010The Double Play
Today SAS Institute had a huge double play of publicity. On the same day, we (1) learned that SAS was rated the No. 1 best place to work in the 2010 Fortune study and (2) announced that SAS achieved record revenues in 2009, despite the recession -- making it 34 straight years of growth. Of course, though the timing of the double play was a coincidence, it is not a coincidence that we are both a great place to work and that our business is doing very well. SAS has always been a great place to work, and SAS has always grown.
I want to highlight four keys to SAS’ success, keys that you may already know: • a culture of change • a culture of customer loyalty • a culture of support • telling our story well SAS has always changed. Most industries need to be kicked into growing into new opportunities, but SAS embraces change. At SAS, we have new initiative areas every year, and adapting to changing needs and conditions has been critical to continue our growth. In their new book Switch, Chip and Dan Heath say that most change management consultants work by introducing an artificial crisis that forces the company to change. The Heath brothers say that this may work in the short term, but that the best change comes with alignment of our elephant and our rider, our resource marshalling and our direction focus. I think that one of SAS' secrets has been in that alignment for change. SAS has always had a culture of customer loyalty. We pioneered an annual renewal model to force us to pay attention to retention and growing our users’ abilities. We have a technical staff base that is well tuned to listen to our customers’ needs and pains. We change in response not just to new opportunities, but in response to current customer needs. SAS has always had a supportive workplace culture, valuing its employees and their needs. We were pioneers in supporting ourselves through day care, through health care, through wellness and fitness programs, and through flexibility, and having a supportive environment not just for ourselves, but for our extended families. We are getting much better at internal communications. We have a nice place to work -- from IT all the way to landscaping. We have it good. SAS is getting good at telling its story. Our category, analytics, is hot, and now we have plenty of listeners to hear our stories in the field of using analytics to solve problems. We lead in a category that is hot. Last summer The New York Times told students to learn statistics if they want a good job. We now have many good books to tell success stories of analytics: Super Crunchers, The Numerati, Competing on Analytics, with more on the way, such as Kaiser Fung's excellent Numbers Will Rule Your World. The public has been trained to be receptive to our message, and we have been getting better at telling our customers’ stories. A key to getting the No. 1 ranking is being able to tell our story well, and the people in corporate communications, internal communications and human resources have done a great job. Congratulations, SAS. You are a great story today, and will continue to be a great story for many years to come. Tuesday, January 19. 2010Science Online 2010 Conference The fourth Science Online conference just wrapped up, and it was as lively as ever. Unfortunately, my attendance was limited since I was coming down with a cold. I attended only a couple of sessions and tried to keep interactions to a minimum, which was hard to do at such an interactive conference with many friendly and familiar faces from previous years. Good thing I've been practicing most of my life at keeping to myself.It was nice to see JMP 8 on the desktop of one of the presentation machines, though I didn't find out whose machine it was. I did make it to Tara Richerson's session on Scientific Visualization, even though it was in a small conference room filled to about three times capacity with most of us on the floor or standing. (The facilities at Sigma Xi RTP were otherwise excellent.) She did a good job leading the discussion, especially considering the broadness of the topic and of the diversity of the audience. Given the amount of interest, visualization could be a separate track onto itself in the future. Tara posited that a good (online) visualization has: Story + Interactivity + Glanceability. Tuesday, December 22. 2009JMP Year-End RoundupAh, yes. It’s been a good year. Why? Because JMP® software only gets better with age. Each new release boasts enhanced features and capabilities. And as time goes on, more people come to recognize the power of data visualization for making breakthroughs. As 2009 comes to a close, we reminisce about the year’s highlights and look to future successes. Amid all the year-in-review coverage you’ll find in newspapers, magazines, and talk radio, here is our very own JMP year-end roundup. Discovery 2009 and Innovators’ Summit
Both events featured Malcolm Gladwell, bestselling author of Blink, Outliers and Tipping Point. Watch clips from his keynote address.
Gladwell discusses a paradigm shift in problem-solving by comparing the circumstances of Watergate, a traditional scandal, to Enron, a modern one. Gladwell suggests that the questions we ask when trying to predict success are too simplistic, even for a seemingly straightforward problem like drafting the best quarterback for a football team. The JMP team looks forward to hosting Discovery Summit 2010, Sept. 13-16 at SAS world headquarters in Cary, NC. Explorers 2009 Seminar Series Users Groups JMP Genomics Social Networking Community 20th Anniversary
And as John Sall, chief architect of JMP, expressed in a video for the occasion, “May JMP always provide enough delights to hold your loyalty.” See you in 2010! Monday, December 21. 2009Which JMP color scheme makes it easier to see water?
Have you ever wanted to color points on a scatter plot by some other variable? I do it often, by using the Row Legend command. This way, not only are the points colored, but a legend is added too.
As an example, let’s create a scatter plot for which the points correspond to zip codes in the continental United States. The x and y coordinates are the longitude and latitude of the zip code, so the scatter plot forms a rough approximation of the country. The data is not current, which is a partial explanation of the sparse data in the western half of the country. Another reason for the sparsity might be that the land areas associated with those zip codes are larger. ![]() Say I want to color the points by another variable: the proportion of the county that is covered with water. I right click on the scatter plot and select Row Legend. The Mark by Column dialog appears. Choose the column which you want to color by. In this case, I want to color by PropWater. ![]() The default JMP color scheme for a continuous variable is called Blue to Gray to Red, and is shown on the picture above. Notice that not only can you color the points according to a column, but you can change the marker as well. The default marker is the regular black dots shown on the scatter plot. For this example, we’ll only color the points. ![]() This color scheme goes from blue (low proportion of water) to gray (middle proportion of water) to red (high proportion of water). The color scheme has various shades of the colors along the spectrum, to account for the fact that Prop Water is a continuous factor. As expected, most of the country is blue. Those counties with more water (the grays and reds) are there, but are hard to isolate and view on the scatter plot. If your purpose is to highlight and focus on those counties with a higher proportion of water, then one of the other JMP color schemes might work better. The other color scheme options can be accessed from the Colors menu on the Mark by Column dialog. Let’s try the White to Blue color scheme. Counties with little water will be closer to the white end of the spectrum, and counties with lots of water will be colored closer to the blue end of the spectrum. That way, the counties with lots of water will stand out, and will be easier to see on the scatter plot. ![]() As you can see, by using a different color scheme, we are able to focus the attention on the counties with lots of water. Notice the blue spot all by its lonesome in the Utah area. That spot didn’t stand out as quickly using the default color scheme. My guess is that the high proportion of water is due to the Great Salt Lake and/or Utah Lake. The color red is often associated with negative/bad values. If a high proportion of water was somehow a bad thing, then you may want to use the White to Red scheme, as shown here. ![]() As we can see from this example, JMP’s color schemes are a powerful asset when visualizing different aspects of your data! Thursday, December 17. 2009JMP® Experiment in ‘Science’
JMP user Peter Reich, a Regents Professor and Distinguished McKnight University Professor at the University of Minnesota, is featured in the December 2009 issue of Science magazine for his long-term experiment on biodiversity and global change. For more than a decade, Reich has used JMP to investigate the interactive effects of elevated levels of carbon dioxide (CO₂) and nitrogen (N) on plant species diversity. Read the customer success story to learn more about that experiment and the ways he analyzes scientific data with JMP.
Perhaps surprisingly, Reich found that although elevated CO₂ reduced diversity by 2 percent and N addition reduced diversity by 16 percent, in combination, N and CO₂ reduced diversity by only 8 percent. “Despite concerns that rising CO₂ and N pollution levels together could lead to as much as 80 percent loss in biodiversity, our 10-year field experiment shows that the joint effects are likely to be much, much smaller,” says Reich in a Science Podcast interview. He is quick to note that while this is good news for biodiversity, the findings don’t lessen the need to curb CO₂ emissions, given the other ways it impacts the planet. Reich and his team conducted the study in synthetic grasslands communities in the Cedar Creek LTER site in Minnesota. Using the Free-Air CO₂ Enrichment Technique (FACE), they manipulated atmospheric conditions, exposing plant species to different combinations of natural or elevated levels of CO₂ and N. The elevated levels simulated future conditions and how ecosystems might react to them. A similar study set in annual serpentine grassland at Jasper Ridge, CA, concludes that, in some settings, climate and atmospheric changes are simply additive combinations – rather than interactive combinations – of each effect by itself. So the question becomes whether or not the results of Reich’s experiment are globally representative. If the CA experiment is more representative, then researchers can “take the knowledge of single global change factors studied one at a time and make models that predict what will happen when many things change at once, and have some confidence that many of these models would be okay,” Reich explains. But more often, predictions made by studying one factor at a time can be inaccurate. With multiple factors (e.g., plants, soil microbes, insects), there is potential for interactions to occur. This makes it more difficult to come up with predictive models that can be applied to other plant communities. Very few government-funded projects have conducted multi-factor experiments that look at the way global change factors such as CO₂, N or climate change influence grasslands, forests and other vegetation. And Reich’s study, now entering its 12th year, is one of the longest-term experiments of its kind. We at JMP like hearing about ways our software is used to expand scientific understanding. If you have a story to tell, please let us know.
Posted by Jessica Marquardt
in Academic, Customer Stories, JMP - General, JMP 8, Statistics, Voice of the Customer (VOC)
at
15:20
| Comments (0)
| Trackbacks (0)
Monday, December 14. 2009New Feature in JMP Genomics 4.1: Tip of the Day
Have you ever wondered if there was a special trick in JMP Genomics for creating a specific graphic or if there was an easy way to change the size of a set of markers?
Well, wonder no more! For the forthcoming release of JMP Genomics, we polled a large group of experts to find out all of the novel ways they use JMP Genomics. We collected their ideas and suggestions into a very cool new feature we call the JMP Genomics Tip of the Day. There are 36 different tips. Each tip shows you a cool trick, explains a feature, or tells you about one or more novel or surprising things you can do with JMP Genomics. A different tip is displayed every time you open JMP Genomics. An example (Tip #35) is shown below. ![]() Suggest a Tip of the Day and get a free JMP Genomics gift Are you a JMP Genomics user? Have you found an interesting feature or shortcut that you would like to share? Just send a description of your tip to genomics@jmp.com. If we include your tip in the next release of JMP Genomics, we’ll send you a free gift. Wait a minute…. I just thought of another one! Tip #37: Although a different tip is surfaced at random each time you open JMP Genomics, you can access all of the tips at any time by clicking Help > Tip of the Day. OK, Shannon, where’s my T-shirt??? Thursday, December 10. 2009It’s in My Blood
We be of one blood, thou and I.
So says Mowgli, the boy raised by wolves in Rudyard Kipling’s The Jungle Book. From Cain and Abel to Twilight’s Edward and Jacob, the human race's very existence has been vitally dependent on that fascinating fluid that flows thicker than water. The ultimate importance of lifeblood has been recognized by science and medicine from their inception, but only relatively recently have some of its deeper technical mysteries been revealed. We’re all familiar with the critical value of a single drop in a forensic case. In Nature Genetics this month, one of our longtime collaborators, Greg Gibson, and his student, Youssef Idaghdour, effectively analyze blood samples from an exotic location to provide some key insights into the mysteries of human genetics and its relationship to the environment and disease. [NOTE: The research project described here had its roots in a 2007 project that is described in a JMP customer story.] Blood nourishes, cleans and defends our primary organs and is an integral component of vital body systems. As one of our most renewable and readily available tissues, it is a very natural target for large-scale genomics studies. An extensive number of clinical trials now routinely collect samples and freeze them away for future investigation. Blood-based biomarkers are highly prized, ranging from basic types (A,B,O were some of the first-used, and various panels are now routinely used in medical labs) to advanced screening of millions of alterations in DNA, RNA, protein, metabolites and other analytes. But there are difficulties: Certain elements of blood fluctuate widely, and some overabundant components like albumin can mask truly discriminating markers. Modern technologies are providing steady progress, and there is even a scientific journal Blood that reports on “bleeding edge” research. Gibson and Idaghdour risked some of their own precious plasma and ventured on camelback into the Moroccan desert to take blood samples from members of two different rural villages. They also collected samples from a nearby urban area. Using bead-based technology from Illumina, they measured more than 500,000 single nucleotide polymorphisms (SNPs, variations in DNA) and 16,000 gene expression levels (RNA transcript abundance) on around 200 individuals. Data analysis was performed in JMP Genomics with help from experts Wendy Czika and Kelci Miclaus in the JMP group. Their findings? Two distinct population ancestries, the Amazigh Berber and Arab, are very apparent in principal component plots of the SNPs, along with individuals in between having an admixture of the two; for a few of them their genetic profiles contradicted their self-reported ancestry! Around a third of the gene expression measurements differ between rural and urban, and 356 are strongly associated with variation in SNPs, so-called expression SNPs (eSNPs). Gender interactions are also apparent, as males tend to travel much more frequently between urban and rural locations than do females. Perhaps most importantly, several of the key eSNPs have direct associations with important diseases like Type 1 diabetes. Gibson is already conducting similar experiments worldwide and is looking to make more extensive connections between genes, environment and disease. While awaiting the next hematological breakthrough, I guess I don’t need to warn you to watch out for viruses, mosquitoes, ticks, leeches, sharks, 18th-century physicians and vampires (any century). You can donate with help from wonderful organizations like the Red Cross, and of course keep tabs on your systolic and diastolic pressure. If you ever find yourself in a desert or jungle and get asked why you like genomics, just reply with a sanguine smile that “It’s in my blood”! Tuesday, December 8. 2009JSL Tip: Multiple Assignments
An often-overlooked feature of the JMP Scripting Language is the ability to assign values to multiple variables in a single assignment statement. A basic example is:
// assign 1 to x and 100 to y The syntax is especially useful when the values of the variables depend on previous values. For instance, an easy (and fast) way to swap the values of two variables is: // swap x and y Here you can see the advantage of the multiple assignment. Without it, you'd have to add a temporary variable and use three assignment statements. Eval List is needed because when JMP evaluates a list, it doesn't evaluate the list members, by default.Another use is when you are performing an iterative calculation on two or more mutually dependent variables. Euclid's greatest common divisor algorithm is one such calculation. // Greatest Common Divisor of two integers Insert a Show(a, b) statement after the assignment to watch how the algorithm works.Wednesday, December 2. 2009Tweeting Live from JMP Seminar in New York
Posted by Arati Bechtel
in Biz Viz, Data Visualization, JMP - General, JMP 8, Tips and Tricks
at
08:30
| Comments (0)
| Trackbacks (0)
Tuesday, December 1. 2009The Twelve Days of Graph Builder
I'm always on the lookout for interesting data. So, when I saw that PNC Wealth Management posted the annual PNC Christmas Price Index, I had to check it out.
PNC tabulates the total current cost of a true love's gifts as enumerated in the carol The Twelve Days of Christmas. This year PNC also provides a handy Excel spreadsheet with all of the historical data back to 1984. I, of course, didn't use Excel to examine it. A quick look at the data in Graph Builder reveals that the seven swans-a-swimming are the most volatile item in the index. ![]() In fact, in the early years they were the most expensive too! PNC recognizes that the volatility may interfere with the index as well so they compute a "Core" index without the swans. Using the Data Filter I left the swans out of Graph Builder as well. ![]() That makes it easier to see that the nine ladies dancing are the most expensive part of Christmas this year. Note to all school-age girls: Learn to dance. It pays much better than milking cows, and it's more fun too. Lastly, as much as I like snowballs, I think that the lines in Graph Builder do a much better job of visualizing this data than the snowballs that PNC uses. ![]() They might look pretty but they don't stand a snowball's chance of being used by serious analysts. Happy holidays! Update: Douglas Okamoto took the PNC data and did some inflationary adjustment and time series analysis. His results are available in the JMP File Exchange.
(Page 1 of 18, totaling 266 entries)
» next page
|
ABOUT THIS BLOG
JMP Statistical Discovery Software from SAS
is proud to bring you this blog on all things related to
data visualization, visual Six Sigma, design of experiments
and other statistical topics.
The blog content appearing on this site does not necessarily represent the opinions of SAS. Your use of this blog is governed by the Terms of Use. CategoriesQuicksearchSyndicate This BlogCONTRIBUTORSCalendar
|
||||||||||||||||||||||||||||||||||||||||||

