turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- JMP User Community
- :
- Discussions
- :
- Discussions
- :
- How to populate additional column with a quantile value

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Sep 16, 2010 10:26 AM
(4538 views)

Hello,

Is there a simple column formula that will populate the quantile of a value in a different column?

For example:

I have a column with the following values [1,2,3,4,5,6,7,8,9,10]. I would like to create a new column that will simply hold the quantile value of each of the values in my original column.

I tried the quantile function in the Column Properties dialogue, but that doesn't seem to be the answer (or I'm using it incorrectly).

Thanks

Is there a simple column formula that will populate the quantile of a value in a different column?

For example:

I have a column with the following values [1,2,3,4,5,6,7,8,9,10]. I would like to create a new column that will simply hold the quantile value of each of the values in my original column.

I tried the quantile function in the Column Properties dialogue, but that doesn't seem to be the answer (or I'm using it incorrectly).

Thanks

2 REPLIES

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

JMP's Quantile functions work the other way around: given a quantile value they produce a data value. If you plot the quantile function (script below), you'll see it has flat spots, which makes it hard to define the inverse function that you want.

You can get close with (Col Rank( : x ) - 1) / (N Row() - 1), but ties are arbitrarily ordered.

The Ranking Tie() function will average ties together, but requires more set-up since it's not a "Col" function.

You can get close with (Col Rank( : x ) - 1) / (N Row() - 1), but ties are arbitrarily ordered.

The Ranking Tie() function will average ties together, but requires more set-up since it's not a "Col" function.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Xan,

Much appreciated. I'll experiment with this. One of my colleagues suggested a workaround which, as it happens, is good enough for what I'm trying to do:

If(:Callback % >= Col Quantile(:Callback %, 0.75), "Upper", "Normal")

Thanks again!

Stefan

Much appreciated. I'll experiment with this. One of my colleagues suggested a workaround which, as it happens, is good enough for what I'm trying to do:

If(:Callback % >= Col Quantile(:Callback %, 0.75), "Upper", "Normal")

Thanks again!

Stefan