Choose Language Hide Translation Bar
Community Member

## "Points Spread" Function: Order of Points

Hello JMP Community,

I am using the Points Spread function in the Fit Y by X platform for oneway analysis. It seems that, by default, the points are spread in reverse order of the order of data in the data table. Meaning, for each X-axis variable, the points are spread from highest row number to lowest row number (left to right).

Why is this, and can I change the order of the points in this platform to match the order in the data table?

Best,

Jenna

1 ACCEPTED SOLUTION

Accepted Solutions
Staff

## Re: "Points Spread" Function: Order of Points

I never noticed how JMP spread the points evenly.

The ordering was likely just a convenient way to create the algorithm to create that spread. After all, the actual position along the x-axis within any group does not matter so any routine that does that is as valid as another. I could imagine it plots from highest row number first because it has to search through the data table to determine the number of observations in a group in order to spread EVENLY. You won't know that result until you hit the end of the table. Just a guess on my part.

Anyhow, one way to get the lowest row number to be plotted furthest to the left and the highest row number plotted furthest to the right, is to re-sort the data table. Create a new column, I will call it Row Number. Initialize it with the default sequence (which will just be a listing from 1 to the number of rows). Now sort your data table in descending order based on Row Number. Now make your plot and spread the points evenly.

I know the points are in reverse order of the data table again, but since you reversed your data table, it works.

Dan Obermiller
Staff

## Re: "Points Spread" Function: Order of Points

I never noticed how JMP spread the points evenly.

The ordering was likely just a convenient way to create the algorithm to create that spread. After all, the actual position along the x-axis within any group does not matter so any routine that does that is as valid as another. I could imagine it plots from highest row number first because it has to search through the data table to determine the number of observations in a group in order to spread EVENLY. You won't know that result until you hit the end of the table. Just a guess on my part.

Anyhow, one way to get the lowest row number to be plotted furthest to the left and the highest row number plotted furthest to the right, is to re-sort the data table. Create a new column, I will call it Row Number. Initialize it with the default sequence (which will just be a listing from 1 to the number of rows). Now sort your data table in descending order based on Row Number. Now make your plot and spread the points evenly.

I know the points are in reverse order of the data table again, but since you reversed your data table, it works.

Dan Obermiller