topic Re: Count how many time Y is greater than N in Discussions

Count how many time Y is greater than N

ark1019 — Thu, 17 May 2018 00:39:41 GMT

Is there a formula that can be used to determine the number of rows that contain a value greater than a predetermined number, like > 65%?

I guess it would be similar to a countif formula in excel. I want to count if the value is greater than 65%.

Thank you

Re: Count how many time Y is greater than N

txnelson — Thu, 17 May 2018 13:09:46 GMT

col sum(if(:your target column > .65, 1, 0));

and this can be shortened to:

col sum(:your target column > .65);

Re: Count how many time Y is greater than N

Jeff_Perkinson — Thu, 17 May 2018 13:46:13 GMT

Of course, if you really only care about knowing the number – as opposed to using it in a further formula – the fastest way to find this is to use the Data Filter (Rows->Data Filter) and then use the filter to select the rows where the value is greater that 65%. You can then see the number of selected rows in the data table.

(view in My Videos)

Re: Count how many time Y is greater than N

tamirlance — Mon, 11 Feb 2019 05:51:29 GMT

Is there a way to do this where 0.65 is not a fixed value but rather the ith value in an adjacent row? Meaning - could I have another cloumn that is a list of percentages and for every row I calculate the sum of the target column greater than the percentage of the ith row of the percentage column?

Something like:

col sum(:your target column > percentage column);

Re: Count how many time Y is greater than N

txnelson — Mon, 11 Feb 2019 06:00:15 GMT

Do you want to only add up the rows where the target column's value is greater than a percentage columns value?

col sum( If(:target column > :percentage column, :target column, 0 );

Re: Count how many time Y is greater than N

tamirlance — Mon, 11 Feb 2019 18:34:57 GMT

yes except I have the complication that the target column is the same as the percentage column. So as I go down the row - I want the value int hat row to act as the threshold for the column

I attached an example data file. I have gotten it to work with summation but it's terribly slow/inefficient and won't work on my real data of >3M rows.

The second column is the summation formula. The third column is the one I want to work. The fourth and fifth columns are variations of the third using fixed values and those evaluate properly for that particular value I put in.

Re: Count how many time Y is greater than N

XanGregg — Tue, 12 Feb 2019 03:38:08 GMT

That sounds a lot like the rank. Col Rank() basically give you the number of items less than a given item (plus 1). For ties, you'd have to use Col Rank(:x, <<tie("minimum")). But you want greater than, so you could subtract the rank from the total number and also substract off the number of equal values:

Col Number( :Binned Temperature )
    - Col Rank( :Binned Temperature, <<tie( "minimum" ))
    - Col Number( :Binned Temperature, :Binned Temperature )
    + 1

Re: Count how many time Y is greater than N

tamirlance — Thu, 14 Feb 2019 00:36:50 GMT

This is great - thanks!