Skip to Main Content

JMP: Working with Data

Exploring data

Reshaping data

JMP Glossary

  • Indicates a continuous variable.
  • Indicates a nominal variable.
  • Sample data tables: Access under Help > Sample Data Library.

JMP Reference

Creating a scatter plot

Analyze > Fit Y by X

Data set-up

You need two continuous variables. This example uses the sample data table Body Fat. (Access through Help > Sample Data Library.)

  • Open the dataset. If you've selected or marked any rows previously and want to start fresh, click on Rows > Clear Row States.

    To create to the scatterplot, click on Analyze > Fit X by Y. This menu will create a scatterplot if you put a continuous variable (small blue triangle) in both the "Y, Response" box and the "X, Factor" box. You can assign variables by clicking on them in the Select Columns list to highlight, then clicking on the "Y, Response" or "X, Factor" buttons. In this example, we make Weight (lbs) the Y and Height (inches) the X. Click OK to make the plot.

  • This will create the scatterplot in a separate window.
  • To added a line fitted to the data, click on the red triangle to the left of the title, "Bivariate Fit of Weight (lbs) By Height (inches)" and select Fit Line.

Descriptive statistics and histograms

Analyze > Distribution

Data set-up

For this example, you need one or more continuous variables. This example uses the sample data table Hurricanes.

  • Open the dataset. If you've selected or marked any rows previously and want to start fresh, click on Rows > Clear Row States.

    Click on Analyze > Distribution. Select the continuous variable you want to describe and move it into the Y, Columns box. (If you select a nominal variable, you will get a frequency table rather than descriptive statistics.) Click OK.

  • You will get a histogram, quantiles, and summary statistics in a new window.

Subsetting data - interactively

Data set-up

These instructions can apply to any dataset. You simply need a variable containing the values that identify which observations to include in your subset. This example uses the sample data table Body Fat.

  • After opening the dataset, the first step is to identify which observations you want to include in your subset. If you've selected or marked any rows previously and want to start fresh, click on Rows > Clear Row States.

    There are different ways to select observations. One way might be to interactively select use a plot such as a scatterplot, which could be a good way to remove outliers. If you follow the instruction to make a scatterplot above, for example, you can select the outliers you want to exclude by holding the Ctrl key and clicking on them, or by drawing a box. The non-selected points will be grayed out, and the number of selected points will display in the lower left of the window.

  • Back in the data table window, you'll see that the selected observations are highlighted with blue, and the Rows box in the lower left, you'll see the total number of rows as well as selected rows.

  • To redo the scatterplot or other analysis with only the selected observation, in the data table window click on Rows > Exclude/Unexclude. You'll see that the Rows box in the lower left will now include a number of excluded observations.

  • Now if you make a scatterplot (Analyze > Fit X by Y), you'll see that the excluded observations won't show up on the plot.

    This is temporary; if you want to include these observations again, just go to Rows > Exclude/Unexclude. If you want to make it permanent, you can use Rows > Delete Rows.

Subsetting data - by group

Tables > Subset

Data set-up

This is useful when you have one data table you'd like to divide into multiple data tables based on the categories of a grouping variables. This example uses the sample data table Hurricanes.

  • In your data table, identify which column contains the grouping variable. Then go to Tables > Subset.

    Make sure to check the box next to Subset By. This will allow you to selecting the grouping variable. Also make sure that "All rows" is selected under Rows, unless you want to include only selected rows. Notice that you can also use this menu to create a random sample of observations.

    Click OK and then you will have a new data table for every value of your grouping variable.

Concatenating data tables

Tables > Concatenate

Data set-up

Concatenating is useful when you want to add rows to a data table. To match up columns, you'll need two or more data tables with identical column names. This example will use sample data tables Trial1 and Trial2.

  • In the window for one data table, click on Table > Concatenate.

    Use the Add button to list the second data table under Data Tables to be Concatenated. If you want to create a new data table leave the box next to "Append to first table" unchecked; to add rows to existing data table, check this box. Check the box next to "Create source column" if you want a column that indicates where each observation came from. Click OK and your data tables will be combined.