COUNTIFS and OR Statement

[thp510]

Does anyone know how to use COUNTIFS and OR statements together? I've only see COUNTIF and OR statements but not COUNTIFS. I'm trying to count the number of times a particular data set has US Citizenship (Y or N), is over 25 years in age, and is either (i.e. OR) born in October 2022, or has a father that was born on October 2022, or has a mother that was born in October 2022.

I have columns for US Citizenship (Y/N), age, birthdate of individual, birthdate of individual's father, and birthdate of individual's mother.

 

[Eric W]

Well, I assume your example is just an example, since otherwise I can tell the answer is 0. No one can be born in October 2022, and be over 25. Given that, there are some complicated ways to use COUNTIFS to incorporate an OR, but you're better off using SUMPRODUCT or FILTER. For example:

Book1
	A	B	C	D	E	F	G	H	I	J	K	L	M
1	Name	Citizen	Age	Birthday	Father Birthday	Mother Birthday			Citizen	Min age	Month	Father Month	Mother month
2	a	Y	23	1/1/2000	10/10/1975	7/7/1979			Y	25	11/1/1990	10/1/1975	3/1/1966
3	b	Y	33	2/2/1990	2/2/1965	3/3/1966
4	c	Y	38	3/3/1985	10/10/1961	10/11/1960			Count
5	d	N	52	10/10/1970	5/5/1949	5/5/1949			2
6	e	Y	32	11/11/1990	1/1/1965	2/2/1966
7	f	Y	23	10/22/1999	10/11/1969	7/24/1969
8	g	N	34	10/1/1988	10/31/1965	8/8/1966
Sheet1

Cell Formulas
Range	Formula
I5	I5	=ROWS(FILTER(A2:A8,(B2:B8=I2)*(C2:C8>=J2)*((EOMONTH(D2:D8+0,0)=EOMONTH(K2,0))+(EOMONTH(E2:E8+0,0)=EOMONTH(L2,0))+(EOMONTH(F2:F8+0,0)=EOMONTH(M2,0)))))
C2:C8	C2	=DATEDIF(D2,TODAY(),"y")

This formula encompasses the basic outline of your example. You can use the same selected date in K2:M2 if you want. Using a formulation like this requires some knowledge of Boolean formulas, which you can research easily.

 

[thp510]

Well, I assume your example is just an example, since otherwise I can tell the answer is 0. No one can be born in October 2022, and be over 25. Given that, there are some complicated ways to use COUNTIFS to incorporate an OR, but you're better off using SUMPRODUCT or FILTER. For example:

Book1
	A	B	C	D	E	F	G	H	I	J	K	L	M
1	Name	Citizen	Age	Birthday	Father Birthday	Mother Birthday			Citizen	Min age	Month	Father Month	Mother month
2	a	Y	23	1/1/2000	10/10/1975	7/7/1979			Y	25	11/1/1990	10/1/1975	3/1/1966
3	b	Y	33	2/2/1990	2/2/1965	3/3/1966
4	c	Y	38	3/3/1985	10/10/1961	10/11/1960			Count
5	d	N	52	10/10/1970	5/5/1949	5/5/1949			2
6	e	Y	32	11/11/1990	1/1/1965	2/2/1966
7	f	Y	23	10/22/1999	10/11/1969	7/24/1969
8	g	N	34	10/1/1988	10/31/1965	8/8/1966
Sheet1

Cell Formulas
Range	Formula
I5	I5	=ROWS(FILTER(A2:A8,(B2:B8=I2)*(C2:C8>=J2)*((EOMONTH(D2:D8+0,0)=EOMONTH(K2,0))+(EOMONTH(E2:E8+0,0)=EOMONTH(L2,0))+(EOMONTH(F2:F8+0,0)=EOMONTH(M2,0)))))
C2:C8	C2	=DATEDIF(D2,TODAY(),"y")

This formula encompasses the basic outline of your example. You can use the same selected date in K2:M2 if you want. Using a formulation like this requires some knowledge of Boolean formulas, which you can research easily.

Thank you @Eric W. Yes, those are dummy examples but I did follow what you did and try to replicate mine to closely match with yours. It looks like I'm able to capture the ORs across the various dates but not the other attributes as it errors out on me (#n/a) when in reality a number should appear. Here's the equation that I have in my google-sheets:

=ROWS(FILTER('Tab1'!$A:$A,('Tab1'!$B:$B=$U17)*((EOMONTH('Tab1'!$N:$N+0,0)=EOMONTH(AV$13,0))+(EOMONTH('Tab1'!$Q:$Q+0,0)=EOMONTH(AV$13,0))+(EOMONTH('Tab1'!$T:$T+0,0)=EOMONTH(AV$13,0)))))

So a few call outs on mine.
- I use full columns in a different raw data worksheet called 'Tab1' versus specifying a very specific set of rows on the same worksheet, as in your example. I'm hoping that's not the reason for my error.
- The part in my formula ('Tab1'!$B:$B=$U17) is checking another formula in column B on the worksheet `Tab1` is equal "Yes". Basically its the (B2:B8=I2) in your equation. When I remove this part of the formula and just check dates, an accurate number shoots out. Not sure why having this formula in the overall function is causing the error.
- I haven't even started adding anything my version of your "(C2:C8>=J2)" part of the equation until I can figure the above out.
- You're probably noticing that I'm also always referencing the AV$13 cell for all my dates. That is because I'm now actually seeing if there's anyone with the same "processing month" (not birthdate). So basically, I want to know how many US Citizens, overage the age of 25, with either their mom, dad, or self having processed through the system in October 2022 (AV$13 = the first day in Oct as 10/1/2022).

Any suggestions?
Tom

 

[Eric W]

I wish you'd mentioned Google Sheets before. There are similarities, but there are differences. Sometimes the tiny differences can make a valid formula in Excel not work in GS, and vice versa. I don't know enough about GS to identify those differences. Still, here are a few comments I have on your questions.

Thank you @Eric W. Yes, those are dummy examples but I did follow what you did and try to replicate mine to closely match with yours. It looks like I'm able to capture the ORs across the various dates but not the other attributes as it errors out on me (#n/a) when in reality a number should appear. Here's the equation that I have in my google-sheets:

=ROWS(FILTER('Tab1'!$A:$A,('Tab1'!$B:$B=$U17)*((EOMONTH('Tab1'!$N:$N+0,0)=EOMONTH(AV$13,0))+(EOMONTH('Tab1'!$Q:$Q+0,0)=EOMONTH(AV$13,0))+(EOMONTH('Tab1'!$T:$T+0,0)=EOMONTH(AV$13,0)))))

This appears OK.

So a few call outs on mine.
- I use full columns in a different raw data worksheet called 'Tab1' versus specifying a very specific set of rows on the same worksheet, as in your example. I'm hoping that's not the reason for my error.

I can't say for sure since it's GS. In Excel using whole column references can be bad since it requires looking at over a million rows, when sometimes far fewer rows will work. I don't know about GS, it might be able to detect the last used row. You can decide if the performance is acceptable with the whole columns. It may be worth experimenting with actual row numbers if you still get an error.

- The part in my formula ('Tab1'!$B:$B=$U17) is checking another formula in column B on the worksheet `Tab1` is equal "Yes". Basically its the (B2:B8=I2) in your equation. When I remove this part of the formula and just check dates, an accurate number shoots out. Not sure why having this formula in the overall function is causing the error.

Hard to say without seeing your sheet. My first guess is that the condition ( ="Yes" ) is not being evaluated as TRUE for some reason. Maybe there's an extra space, maybe capitalization counts, maybe you're looking at the wrong column. If GS has the equivalent of "Evaluate Formula", you can check with that. You can also try ISNUMBER(SEARCH("Y",'Tab1'!$B:$B)) if GS supports it, to just look for the letter Y.

- I haven't even started adding anything my version of your "(C2:C8>=J2)" part of the equation until I can figure the above out.

Logical, get one issue resolved before starting another.

- You're probably noticing that I'm also always referencing the AV$13 cell for all my dates. That is because I'm now actually seeing if there's anyone with the same "processing month" (not birthdate). So basically, I want to know how many US Citizens, overage the age of 25, with either their mom, dad, or self having processed through the system in October 2022 (AV$13 = the first day in Oct as 10/1/2022).

Processing month makes more sense than birthdate, I wondered about that.

Any suggestions?
Tom

You seem to have the basic idea, just try to work out the issues. Good luck!

 

[thp510]

I wish you'd mentioned Google Sheets before. There are similarities, but there are differences. Sometimes the tiny differences can make a valid formula in Excel not work in GS, and vice versa. I don't know enough about GS to identify those differences. Still, here are a few comments I have on your questions.


This appears OK.

I can't say for sure since it's GS. In Excel using whole column references can be bad since it requires looking at over a million rows, when sometimes far fewer rows will work. I don't know about GS, it might be able to detect the last used row. You can decide if the performance is acceptable with the whole columns. It may be worth experimenting with actual row numbers if you still get an error.

Hard to say without seeing your sheet. My first guess is that the condition ( ="Yes" ) is not being evaluated as TRUE for some reason. Maybe there's an extra space, maybe capitalization counts, maybe you're looking at the wrong column. If GS has the equivalent of "Evaluate Formula", you can check with that. You can also try ISNUMBER(SEARCH("Y",'Tab1'!$B:$B)) if GS supports it, to just look for the letter Y.

Logical, get one issue resolved before starting another.

Processing month makes more sense than birthdate, I wondered about that.

You seem to have the basic idea, just try to work out the issues. Good luck!

Thank you again.