Subscribe Bookmark
arati_mejdal

Staff

Joined:

May 21, 2014

Analytics and daily life

JMP software is used for making important and compelling business points and scientific discoveries. Having worked in the software industry for a number of years, I’ve also found an interesting personal reliance on JMP to answer simpler questions in my daily life. I realized this recently as I maneuvered through my yearly personal budgeting process and shopped for a new place to live. What is it that’s said about having a hammer and everything looking like a nail?

For personal budgeting, I sometimes use a website called Mint.com. This website pulls data from banks, loans and credit cards into one Web source, and it allows me to export data into one .csv file.  Clicking on the .csv file automatically launches Excel on my system, which has a handy JMP add-in available to export data directly from Excel to JMP. One of my favorite aspects of JMP is the ease of importing data from other sources like Excel, websites with tables and .csv files.

Using the JMP Add-In for Excel, with one button click I create a JMP data table and then start working to unravel the  stories in my spending habits. Mint’s data format groups my spending into categories that make it easy for me to ask questions in JMP. Am I eating out too much? Should I buy more groceries? What’s the impact of my visit to the coffee shop every morning?

 

Like most JMP users, I have a pattern in the way I work with JMP to navigate my way through an initial data session. My current approach uses Graph Builder as the first step, adding the Local Data Filter, which became available in JMP 10, to test different theories interactively. The Local Data Filter can be combined with a platform window for ad-hoc subsetting without affecting row states in a particular platform analysis. Attaching it to my Graph Builder session allows me to make an immediate impact on my view with selection criteria that are specifically targeted to my questions and theories.

In JMP 10, we also introduced a coxcomb graph into the Graph Builder collection. In simple cases, this graph can be a compelling tool to show highlights in data that would typically fit a pie or bar chart analysis.

The coxcomb graph varies the radius on a traditional pie chart slice to reflect summary statistics. Thus, in my simple budgeting case, this graph shows me that I might do better to buy more groceries.

 

 

I also used Graph Builder with Wake County (North Carolina, USA) real estate data provided by a local broker, to clearly see how long I might expect my house to stay on the market. Looking at the graph below, I can see that as long as I price my home within market expectations, I should be fine. Most houses less than $700,000 were sold for over 95% of their asking price. Homes costing $1 million or more were sold for a much smaller percentage of the asking price and took many more days to sell. I guess that folks who own million-dollar homes can afford to hold off dropping their price.

A coxcomb graph can show more levels of information within each pie slice. Continuing with my analysis of daily habits, I’ll give some thought to the health implications of my choices in candy bars, using JMP’s sample data table candybars.jmp. In this data table, there is an assortment of candy bars, along with information about their calorie content along with their healthier characteristics, like vitamins and fiber.

Using the Local Data Filter, I again make my selections and review my top contenders.

Not surprisingly, I immediately see that the calorie count is much higher for my favorite candy bar. Luckily, this graph also shows me that the fiber and vitamins in this candy bar are also higher, thus making it a little easier to leave these on my grocery list as I improve my budgeting and eating habits in 2012!

 

1 Comment
Community Member

Audrey Ventura wrote:

I love it - using JMP to rationalize the particular candy bar purchase!