Subtraction in Column Formula Slow

[#1]

Hello!

If I use analyze>tabulate>mean, then jmp 14 can take my 3 million rows and get a summary in about 15 seconds. 

If I try to subtract column A from column B using a column formula, it takes 3+ hours. 

How can I do a workaround to increase calculation speed? some sort of JSL? 

I could just pull it into python but jmp should be able to handle subtraction! 

Thanks for your help! 

[#2]

Running Windows 11 and JMP 17.2, I created a data table with 2 columns of 3,000,000 rows and then added a new column with the formula

 - :a

and the calculation of the formula row is virtually instantaneous.

Can you provide more specifics of your issue that might provide some insights? 

[#3]

I am also running Windows 11 and JMP 17.2 and I used a data file with 6 million rows and both tabulate and the formula column calculate almost instantly (1 second).  I suspect it is your hardware - on my older machines, things used to take much longer.  15 seconds sounds too long for the tabulate.  But I don't know if it is hardware, JMP 14 (which I no longer have to test), or a combination of the two.

[#4]

I ran the same test on JMP 14 and while it was slightly slower, it did the subtraction in less than a second

[#5 · Accepted solution]

Are there platforms (distribution, etc) open or subsets linked to the table while the formula is running? Possibly the platforms are updating as the formula progresses.

If you are scripting, dt<<runformulas will run in the foreground (faster).

[#6]

Thanks Craig and everyone else for your input. 

I did have a linked table open. Closed it and the math finished instantly, as expected!