Visualizing and Exploring Data

[#1]

After your data is prepared for analysis, use JMP to interactively explore, uncover hidden relationships, and start understanding your data. Learn how to uncover relationships visually to identify and correct missing information, review key summary statistics, and examine subsets of interest uncovered during exploration.

See how to:

• Interpret column header graphss
• Get the most from dyynamic linking
• Recode one or more values easily and interactively
• Use Columns Viewer
• Detect and correct  messy source information
• Explore using Graph Builder and Control Chart Builder
• Arrange values in the sequence of your choice
• Filter data interactively

Questions answered by Yasmine Hajar @yasmine_hajar and Greg Flexman @GregF_JMP  at the live webinar:

Q: Is there a way to "undo" back to your previous selection? Like if  I manually select a bunch of rows, then accidentally click on a column header and lose my selection, is there a way to return to my previous selection?

A: No. That's because "selection" is not considered a "Step".  (Steps are saved to the log, in case you want to check to see if an action is saved.) But you can definitely "undo" creating a column for example.

Q: What is the subgroup (x-axis on the control chart)?

A: In the case you have let's say batch NBS, you can drag that column to the subgroup axis and JMP will show you the average value of the parameter per group rather than every single data point.

Q: What if you don't define the subgroup, like was done in this example?

A: Nothing bad will happen. Your x axis is ordered by the rows. Each row is a new data point. Not to get into too many details, but the Control Chart is called an Individual chart when no subgroups are assigned and Xbar chart when you have subgroups. You can look them up in the documentation if you are interested.

Q: What is the meaning behind the numbers being generated on the x axis if you don't define the subgroup?

A: Number of Row(s)

Q: Are you Right clicking the graph builder columns to add the different grouping?

A: it's a drag and drop of the columns to the X or Y axis. You can add multiple columns in the same time as well.  The right click and left click serve different default functions. (But, remember, some people customize their mouse buttons.) In general, right click to drag and drop into the graph area, and left click from the graph area to get options for that part of the graph.  See how:

Q: How did you save it so Accept is blue and Reject is yellow?

A: To change the colors, you can use Row State.  lso, you can also create another column to solve Color and any other (like HIDE/Exclude) Row States. See how:

Bonus: Here are some related answers from early webinars on this topic:

Q: How do I handle spec limits that are different for different rows of data?

A: See this JMP Community article by a JMP Developer:  Process specification limits that differ based on a grouping column . It goes over how to input multiple spec limits that differ based on a grouping column (ex: Supplier Name). It will then separate the capability analysis based off the grouping and the differing spec limits.

Q: How do you start recoding, like you did for Jones and Smith?

A: You can access the recode utility by right-clicking on the column name in the data table or from the Cols menu at the top.

Q:  How do you have the LSL (Lower Specification Limit) line shown?

A: Spec limits are handled by column properties in JMP. When you specify the spec limit column property, there is a check box option to show reference lines for the defined specs.

Q: Can we do caption box plots?

A: You can add captions by going to the Tools menu and select Annotate. You can use this to add captions to any JMP report.

Q: Is there a way to show spec lines by default on graphs? I have to manually do it each time.

A:  You can set a JMP Preference for this. Go to File > Preferences > Platforms > Manage Limits, and check the "Show Limits All" option.

Q: Explain what the normal quantile chart means please? Why is this important?

A: One use is it allows you to determine if your distribution is normally distributed if the normal quantile chart shows a linear relationship.

Q: How do you access column view selector?     

A: You can access column viewer from the Column menu.

Q: Does "recode" create a new column?

A: You can have it create a new column or have it recode in the existing column.

Q: When do I see Distribution Frequencies instead of Quantiles and Summary Statistics?

A: Frequencies are shown for columns with categorical modeling types, either nominal (red bars) or ordinal (green bars).  Continuous data (blue triangles) shows the quantiles and summary statistics.

Q: If replacing the existing column during recode, can i reverse it back later?

A: You can undo the recode to get back the existing values.  In general, it's good practice to make a new column so that you preserve the original values.

Q: Does LOCAL data filter mean it's only filtering in the window and not doing anything to the underlying data table?

A: Yes               

Q: Is there a way to keep range of graphs fixed during filter animation?

A: Yes, from the red triangle menu in graph builder, select the "Lock Scales" option.

Q: Can we capture Animation effect (in Local Data Filter) in a PowerPoint?           

A: In the windows version of JMP you can.  Select animation from the red triangle of the data filter.  Click the record button in the Local Data Filter, turn on the filter animation, then click stop and then save the result as a .gif file.  If you insert that .gif into PPT, it will animate when you present the slide.

Accessing Animation From Data Filter Red Triangle

Q: How can I make a graph with x-axis showing # of observations, y-axis showing accumulative # of sale representatives?

A: In addition to Nick's answer, there is a Cumulative Sum summary statistic available in Graph Builder when you make a Bar Chart.

Q: Can you display the number of data points for box plot on the Graph Builder Graph?

A: When you do a Oneway analysis from the Fit Y by X platform and turn on the "Means and Std Dev" option, you will see the number of observations in each categorical level.

Q: Can we group scripts on JMP data table for common categories?

A: Yes. Select the scripts you wish to group, right click and select Group Scripts.

Resources

• Free self-paced JMP course on Statistical Thinking for Industrial Problem Solving
• NEW JMP self-paced Statistical Process Control course.
• More about Workflow Builder (includes video)

[#2]

See how to:

• Interpret column header graphss
• Get the most from dyynamic linking
• Recode one or more values easily and interactively
• Use Columns Viewer
• Detect and correct  messy source information
• Explore using Graph Builder and Control Chart Builder
• Arrange values in the sequence of your choice
• Filter data interactively

Questions answered by Yasmine Hajar @yasmine_hajar and Greg Flexman @GregF_JMP  at the live webinar:

Q: Is there a way to "undo" back to your previous selection? Like if  I manually select a bunch of rows, then accidentally click on a column header and lose my selection, is there a way to return to my previous selection?

A: No. That's because "selection" is not considered a "Step".  (Steps are saved to the log, in case you want to check to see if an action is saved.) But you can definitely "undo" creating a column for example.

Q: What is the subgroup (x-axis on the control chart)?

A: In the case you have let's say batch NBS, you can drag that column to the subgroup axis and JMP will show you the average value of the parameter per group rather than every single data point.

Q: What if you don't define the subgroup, like was done in this example?

A: Nothing bad will happen. Your x axis is ordered by the rows. Each row is a new data point. Not to get into too many details, but the Control Chart is called an Individual chart when no subgroups are assigned and Xbar chart when you have subgroups. You can look them up in the documentation if you are interested.

Q: What is the meaning behind the numbers being generated on the x axis if you don't define the subgroup?

A: Number of Row(s)

Q: Are you Right clicking the graph builder columns to add the different grouping?

A: it's a drag and drop of the columns to the X or Y axis. You can add multiple columns in the same time as well.  The right click and left click serve different default functions. (But, remember, some people customize their mouse buttons.) In general, right click to drag and drop into the graph area, and left click from the graph area to get options for that part of the graph.  See how:

Q: How did you save it so Accept is blue and Reject is yellow?

A: To change the colors, you can use Row State.  lso, you can also create another column to solve Color and any other (like HIDE/Exclude) Row States. See how:

Bonus: Here are some related answers from early webinars on this topic:

Q: How do I handle spec limits that are different for different rows of data?

A: See this JMP Community article by a JMP Developer:  Process specification limits that differ based on a grouping column . It goes over how to input multiple spec limits that differ based on a grouping column (ex: Supplier Name). It will then separate the capability analysis based off the grouping and the differing spec limits.

Q: How do you start recoding, like you did for Jones and Smith?

A: You can access the recode utility by right-clicking on the column name in the data table or from the Cols menu at the top.

Q:  How do you have the LSL (Lower Specification Limit) line shown?

A: Spec limits are handled by column properties in JMP. When you specify the spec limit column property, there is a check box option to show reference lines for the defined specs.

Q: Can we do caption box plots?

A: You can add captions by going to the Tools menu and select Annotate. You can use this to add captions to any JMP report.

Q: Is there a way to show spec lines by default on graphs? I have to manually do it each time.

A:  You can set a JMP Preference for this. Go to File > Preferences > Platforms > Manage Limits, and check the "Show Limits All" option.

Q: Explain what the normal quantile chart means please? Why is this important?

A: One use is it allows you to determine if your distribution is normally distributed if the normal quantile chart shows a linear relationship.

Q: How do you access column view selector?     

A: You can access column viewer from the Column menu.

Q: Does "recode" create a new column?

A: You can have it create a new column or have it recode in the existing column.

Q: When do I see Distribution Frequencies instead of Quantiles and Summary Statistics?

A: Frequencies are shown for columns with categorical modeling types, either nominal (red bars) or ordinal (green bars).  Continuous data (blue triangles) shows the quantiles and summary statistics.

Q: If replacing the existing column during recode, can i reverse it back later?

A: You can undo the recode to get back the existing values.  In general, it's good practice to make a new column so that you preserve the original values.

Q: Does LOCAL data filter mean it's only filtering in the window and not doing anything to the underlying data table?

A: Yes               

Q: Is there a way to keep range of graphs fixed during filter animation?

A: Yes, from the red triangle menu in graph builder, select the "Lock Scales" option.

Q: Can we capture Animation effect (in Local Data Filter) in a PowerPoint?           

A: In the windows version of JMP you can.  Select animation from the red triangle of the data filter.  Click the record button in the Local Data Filter, turn on the filter animation, then click stop and then save the result as a .gif file.  If you insert that .gif into PPT, it will animate when you present the slide.

Accessing Animation From Data Filter Red Triangle

Q: How can I make a graph with x-axis showing # of observations, y-axis showing accumulative # of sale representatives?

A: In addition to Nick's answer, there is a Cumulative Sum summary statistic available in Graph Builder when you make a Bar Chart.

Q: Can you display the number of data points for box plot on the Graph Builder Graph?

A: When you do a Oneway analysis from the Fit Y by X platform and turn on the "Means and Std Dev" option, you will see the number of observations in each categorical level.

Q: Can we group scripts on JMP data table for common categories?

A: Yes. Select the scripts you wish to group, right click and select Group Scripts.

Resources

• Free self-paced JMP course on Statistical Thinking for Industrial Problem Solving
• NEW JMP self-paced Statistical Process Control course.
• More about Workflow Builder (includes video)