Google Sheets : horizontal (row) conditional Arrayformula - google-sheets

I have row 1 containing some dates (undetermined number, because end formula will be a dynamic number) and I want row 5 to display the average of every column that has a date as a header.
Trying to have this array formula work :
arrayformula(iferror(
if(B1:1 = "",,
AVERAGE(
indirect(
SUBSTITUTE(ADDRESS(1,COLUMN(B1:1),4),"1","")
& 7 & ":" &
SUBSTITUTE(ADDRESS(1,COLUMN(B1:1),4),"1","")
)))))
I think I did isolate the problem in the arrayformula, I can't see where the problem lies with it ?
Here's a screenshot :
and here's the link to the spreadsheet :
https://docs.google.com/spreadsheets/d/1GbIWwv0lq6nVdn4dHkSO0-QqZYEPCwJtntCu8wcSd9Q/edit?usp=sharing

QUERY method:
=ArrayFormula(IF(B1:1="",,IFERROR(VLOOKUP(COLUMN(B1:1),QUERY(SPLIT(FLATTEN(TRANSPOSE(COLUMN(B7:7)&"♦"&B7:1000)),"♦"),"select Col1,avg(Col2) where Col2 is not null group by Col1 label avg(Col2) ''",0),2,FALSE),)))
MMULT method:
=ArrayFormula(IF(B1:1="",,IFERROR(MMULT(N(TRANSPOSE(ROW(B7:1000)^0)),N(B7:1000))/MMULT(N(TRANSPOSE(ROW(B7:1000)^0)),N(B7:1000<>"")),)))

Related

Count Values under Columns with Same Name Google Sheets

I have a Google sheet which has columns with the same name and there are different values under each column. I want to count the same value that appear under the same column name.
1
2
3
1
2
R
B
C
R
D
D
C
R
B
D
For example, I would like to get the number "R" that appear under column "1", so I would expect a count of 2 for "R" appearing under columns 1.
Here is a link to Google Sheet with actual data.
I have tried countif and countifs in Google Sheets, but can't figure out how to get the count right based on column name.
Try this formula, it outputs an array which shows how many of each letters are contained in each column name:
=LAMBDA(NUMBERS,LETTERS,
LAMBDA(UNUM,ULET,
{
{"",TRANSPOSE(UNUM)};
{ULET,
MAKEARRAY(COUNTA(ULET),COUNTA(UNUM),LAMBDA(ROW,COL,
COUNTIF(FILTER(LETTERS,NUMBERS=INDEX(UNUM,COL)),INDEX(ULET,ROW))
))
}
}
)(UNIQUE(FLATTEN(NUMBERS)),UNIQUE(FLATTEN(LETTERS)))
)($A$1:$AE$1,$A$2:$AE$18)
Assume that your sample datarange is A1:AE18.
apply UNIQUE() and FLATTEN() to A1:AE1, to get the unique entries of column names.
apply UNIQUE() and FLATTEN() to A2:AE18, to get the unique entries of data.
use LAMBDA() to name the dataranges and output of step 1 & 2 as:
NUMBERS (=A1:AE1),
LETTERS (=A2:AE18),
UNUM (=UNIQUE(FLATTEN(NUMBERS))),
ULET (=UNIQUE(FLATTEN(LETTERS))).
create Arrays with {}, which...
1st column's value is a blank, followed by TRANSPOSE(UNUM) in the row,
1st row's value is a blank, followed by ULET in the column.
inside the above said range, use MAKEARRAY() to create results.
MAKEARRAY() set an array by defining the length of ROW and COL, which we uses...
COUNTA(ULET) as the number of rows and,
COUNTA(UNUM) as the number of columns.
inside MAKEARRAY(), you also need a LAMBDA() to apply what to do with each CELL of the new created array, each CELL is accessed by the ROW and COL index.
in our case, we set up the row and col number of the new array using ULET and UNUM. Therefor, the index of each CELL of the new array will be equal to the index of each value inside ULET and UNUM, we can than take that as reference and use COUNTIF() with FILTER() to calculate the number of repeats of each letter in each column name.
You can try this:
= ARRAYFORMULA(
query(
query(
SPLIT(TRANSPOSE(SPLIT(
QUERY(
TRANSPOSE(
QUERY(
TRANSPOSE(
IF(Original!A2:AE18<>"",
"😊"&Original!A1:AE1&"♥"&Original!A2:AE18, )
),,999^99)
),,999^99),
"😊")),
"♥"),
"Select Col1,Col2,count(Col2) group by Col1,Col2"),
"Select max(Col2),Col3 group by Col2,Col3 pivot Col1")
)
Note: (Got inspired by player0 useful answers)
Output:
We can read from the table that: R is appearing 40 times under the column named '1', 24 times under the colum named '2', etc...
The following is a more compact approach than the previous answers:
=arrayformula(query(split(flatten(A1:AE1&"|"&A2:AE18),"|"),"select Col2,count(Col2) group by Col2 pivot Col1"))
N.B. I'm assuming the order of the grouped values in each column is irrelevant, so the QUERY default of lexicographical ordering is fine.

Counting over aggregated columns in Google Sheets

I have the yellow table shown below, and I'm trying to get the blue table, which aggregates columns B:F by value, and then counts the number of 'x' symbols for each row value of column A.
Is there some basic SQL/array magic formula to get this, please? There must be.
Use this new functions formula
=BYROW(B2:4, LAMBDA(v, COUNTIF(v, "=x")))
Used:
BYROW, LAMBDA, COUNTIF
v is the array_or_range
Update
={ A2:A4, BYROW(B2:4, LAMBDA(vv, COUNTIF(vv, "=x")))}
For fun
Update 02
=ArrayFormula(TRANSPOSE(QUERY({
QUERY(TRANSPOSE(IF(A1:4<>"x",A1:4,1)),
" Select * Where Col1 is not null ", 1)},
" Select (Col1),sum(Col2),sum(Col3),sum(Col4) Group by Col1 ", 1)))

Query and Importrange to get the last value of the selected row of spreadsheet | Google Sheets

Right now I am using this query to search for a row based on its Column 1 value. Then it takes the value from the last column. I need a way for it to automatically find the last column in the row since some of the rows have more columns than others.
This is what I had before, which I had manually specified the last column with a value:
=QUERY(IMPORTRANGE("link_redacted","PriceList!A1:AZ100000"), "Select Col10 where Col1 = '5531001'",1)
I have tried using LOOKUP with ARRAYFORMULA I couldn't get it to work:
=QUERY(IMPORTRANGE("link_redacted","PriceList!A1:AZ100000"), "Select (LOOKUP(1, ARRAYFORMULA(1/[Select Col1 where Col1 = '5531006']:[Select Col100 where Col1 = '5531006']<>"")[Select Col1 where Col1 = '5531006']:[Select Col100 where Col1 = '5531006']))",1)
Any ideas for a simpler way to do this?
Since no example is presented, I tested the formulas given but no source data is fetched.
so i created a a minimal, reproducible example
Example is the data on the left
Use this formula to get the last non empty columns values.
=ArrayFormula(IFERROR( REGEXEXTRACT( TRIM(TRANSPOSE(QUERY(TRANSPOSE(C3:E),,ROW(C3:E)))), "[^\s]+$")))

GoogleSheets SumIFS in array formula

I'm trying to convert my Sumifs formula into an array formula
(so that the sheet auto updates)
I'm struggling to do so though
I tried using sumif formula and combining the columns, but I can't get it to work right.
I created this simplified version of my data
https://docs.google.com/spreadsheets/d/1qnoY_MZTup8XLLWCu61-Ifiv8dak6QRc1CxsGKpmFWE/edit?usp=sharing
(I know there's only two rows, but in my actual data there's a lot more and the number of brands vary)
I've shown in column C on sheet 2 how the sumifs formula works.
But ideally I'd like an array formula in cell B2 that will populate all the rows for me (instead of having to fill down the sumifs each time)
No need to use arrayformula for this. You can simply use query:
Formula (A2):
=query({Sheet1!A2:D}, "select Col1, sum(Col4) where Col1 is not null and Col2 is null and Col3 is null group by Col1 label sum(Col4) ''")
This will populate both A and B.
Output:

Google Sheet | Excel | Array Formula + CountIf + Partial Text Problem

I'm pretty new with ArrayFormula, have been trying but sometime the formula works, sometimes does not. What I'm trying to do is the combination of ArrayFormula, Countif for searching partial text.
As shown in the worksheet below, there are 10 subjects (column A), each subject has at least one of 4 samples (A,B,C,D) summarized as a string (column B). What I'm trying to do is to find which subject has sample A or B or C or D.
I have tried single formula for each sample, eg cell D3
=IF(COUNTIF($B3,"*"&$D$2&"*")>0,$A3,"")
it returns the correct results. However, when I try arrayformula in cell I3,
=arrayformula(IF(COUNTIF($B3:B,"*"&$D$2&"*")>0,$A3:A,""))
The answers are weird. For example: Subjects (Gamma, Zeta, Eta, Theta) who don't have the sample "A" are shown to have sample "A". And this applies to sample B,C,D too
Not sure what went wrong in here. Here is the link to the worksheet
I wouldn't use Countifs or an array formula. Use filter instead. Put this formula in cell i3.
=Filter(if(REGEXMATCH(B3:B,$D$2),A3:A,""),B3:B<>"")
try:
=INDEX(QUERY(IFERROR(TRIM(SPLIT(FLATTEN(IF(IFERROR(SPLIT(B3:B, ","))="",,
SPLIT(B3:B, ",")&"×"&A3:A)), "×"))),
"select max(Col2) where Col2 is not null group by Col2 pivot Col1"))
or use in row 2 if you want to sort it as in your example:
=INDEX(IFNA(VLOOKUP(A2:A, QUERY(IFERROR(TRIM(SPLIT(FLATTEN(
IF(IFERROR(SPLIT(B3:B, ","))="",,SPLIT(B3:B, ",")&"×"&A3:A)), "×"))),
"select Col2,max(Col2) where Col2 is not null group by Col2
pivot Col1 label Col2'Subjects'"), {2,3,4,5}, 0)))
You can accomplish all four columns of results with a single formula.
Delete all formulas from I3:L3.
Place the following formula into I3:
=ArrayFormula(IF(REGEXMATCH(B3:B,I2:L2),A3:A,))
In plain speech, this read "If anything in B3:B matches a value found in I2:L2, return A3:A in the matching columns(s) at the matching row(s); if not, return null."

Resources