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
- :
- help creating a column that is a cumulative sum based on other columns

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

Aug 17, 2016 11:51 AM
(5456 views)

I am trying to create a new column in JSL that will be an ongoing sum of the previous rows in that column. I have a column called "Total Daily Flow" I also have a column which marks what location the data in "total daily flow came from (call that "Site"). for example, Say on Oct. 1, (which is specified in another column called "day of water year") i have a total daily flow of 1000, and this day happens to be day 1 of the year. I want a column whose first row will be = 1000. then say on day two, which will be the second row in the table, Oct. 2, there is a total daily flow of 5000, I would want the new row in the new column to be 1000 + 5000 = 6000. then say Oct. 3 has a daily flow of 3000, so the next row will be 6000+3000 = 9000. This is to go on for every row in my table that shares the same location name (Site) and is in the same "year" as specified by the column "day of water year" (1-365).

2 ACCEPTED SOLUTIONS

Accepted Solutions

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

Aug 17, 2016 1:31 PM
(8389 views)

hi aporter,

until someone smarter comes up with a one liner you can try my plumbers solution:

Names Default To Here**(** **1** **)**;

dt = Open**(** "$SAMPLE_DATA/Big Class.jmp" **)**;

dt << **New Column(** "Row", formula**(** :Row**()** **)**, eval formula **)**;

dt:row << **delete formula**;

// assuming age is time and sex is site

dt << **Sort(** By**(** :sex, :age **)**, Order**(** Ascending, Ascending **)**, replace table **)**;

// now sum as you want it

dt << **New Column(** "sumhightbysex",

formula**(** If**(** Row**()** == **1**, :height, :sex == Lag**(** :sex, **1** **)**, :height + Lag**(** :height, **1** **)**, :sex != Lag**(** :sex, **1** **)**, :height, Empty**()** **)** **)**, eval formula

**)**;

dt:sumhightbysex << **delete formula**;

// sort to original order

dt << **Sort(** By**(** :Row **)**, Order**(** Ascending **)**, replace table **)**;

// clean up

dt << **delete column** **(**row**)**;

Speak**(** "bob's your uncle" **)**;

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

Sep 8, 2016 1:11 PM
(8388 views)

Hi Ron,

This is a great solution, thanks for posting.

The formula for the new column sumheightbysex has a slight typo. It should be

If(Row() == 1, :height, :sex == Lag(:sex, 1), :height + Lag(:sumhightbysex, 1), :sex != Lag(:sex, 1), :height, Empty())

using the value in the new column to add to the height, and not the height from previous row, :height + Lag(:height), 1).

G'day

4 REPLIES

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

Aug 17, 2016 1:18 PM
(5284 views)

If you are running JMP version 12, if you right click on the column header field in the data table, the New Formula Column -> Row -> Cumulative should work without having to resort to JSL. See below:

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

Aug 17, 2016 1:31 PM
(8390 views)

hi aporter,

until someone smarter comes up with a one liner you can try my plumbers solution:

Names Default To Here**(** **1** **)**;

dt = Open**(** "$SAMPLE_DATA/Big Class.jmp" **)**;

dt << **New Column(** "Row", formula**(** :Row**()** **)**, eval formula **)**;

dt:row << **delete formula**;

// assuming age is time and sex is site

dt << **Sort(** By**(** :sex, :age **)**, Order**(** Ascending, Ascending **)**, replace table **)**;

// now sum as you want it

dt << **New Column(** "sumhightbysex",

formula**(** If**(** Row**()** == **1**, :height, :sex == Lag**(** :sex, **1** **)**, :height + Lag**(** :height, **1** **)**, :sex != Lag**(** :sex, **1** **)**, :height, Empty**()** **)** **)**, eval formula

**)**;

dt:sumhightbysex << **delete formula**;

// sort to original order

dt << **Sort(** By**(** :Row **)**, Order**(** Ascending **)**, replace table **)**;

// clean up

dt << **delete column** **(**row**)**;

Speak**(** "bob's your uncle" **)**;

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

Sep 8, 2016 1:11 PM
(8389 views)

Hi Ron,

This is a great solution, thanks for posting.

The formula for the new column sumheightbysex has a slight typo. It should be

If(Row() == 1, :height, :sex == Lag(:sex, 1), :height + Lag(:sumhightbysex, 1), :sex != Lag(:sex, 1), :height, Empty())

using the value in the new column to add to the height, and not the height from previous row, :height + Lag(:height), 1).

G'day

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

Sep 12, 2016 9:10 AM
(5284 views)

glad you find it useful,

Perhaps the oneliner can be based on this:

Re: Syntax: How do I create an index column by ID

it would also make it independent of sorting and maintain an active formula.

but i am not sure how to tackle it.

Best