I have the below table, and have created the Wk2-Wk3 column by using a nested IF/AND function
=
IF(AND(C3="Newbie",D3="Promise"),"N-P",
IF(AND(C3="Newbie",D3="Sleepy"),"N-S",
...
IF(AND(C3="Broken Promise",D3="Newbie"),"Bp-N",
IF(AND(C3="Fallen Star",D3="Newbie"),"Fs-N"
)))))))))))))
ID
Wk 1
Wk 2
Wk 3
Wk2-Wk3
1
Newbie
Promise
Waning Promise
P-Wp
2
Newbie
Sleepy
Hibernating
S-H
3
Promise
Promise
Star
P-S
4
Promise
Waning Promise
Broken Promise
Wp-Bp
5
Hibernating
Hibernating
Newbie
H-N
6
Newbie
Newbie
Promise
N-P
However, as the weeks progress I will be adding additional 'Wk' columns and therefore the last column will be changing to Wk3-Wk4, Wk4-Wk5,
.. and so on.
I don't want to have to keep altering cell references. Is there a way to find the last column using a fixed formula that always uses the two latest weeks.
I tried using OFFSET, but cannot find a way to reference in relation to the formula cell.
The following would not work:
=
IF(AND(((OFFSET(E2,0,-2))="Newbie",((OFFSET(E2,0,-1))="Promise"),"N-P",
IF(AND(((OFFSET(E2,0,-2))="Newbie",((OFFSET(E2,0,-1))="Sleepy"),"N-S",
...
IF(AND(((OFFSET(E2,0,-2))="Broken Promise",((OFFSET(E2,0,-1))="Newbie"),"Bp-N",
IF(AND(((OFFSET(E2,0,-2))="Fallen Star",((OFFSET(E2,0,-1))="Newbie"),"Fs-N"
)))))))))))))
Using latest version of Excel, thank you in advance
Alternatively use INDEX() to retrieve the two cells to the left of the cell holding the formula:
Formula in E2:
=REDUCE("",INDEX(2:2,COLUMN()-{2,1}),LAMBDA(a,b,TEXTJOIN("-",,a,PROPER(CONCAT(LEFT(TEXTSPLIT(b," ")))))))
Looking at the logic in your IF() it seems you don't really need all these nested statements if you actually just need the first letter of each word. I did mimic this logic and used TEXTSPLIT() to retrieve every leftmost character of each word in the cells, concatenate them and use PROPER() before TEXTJOIN().
This would, however, require access to ms365's insiders channel.
If I understand correctly, then OFFSET is indeed what you can use, it just might seem momentarily confusing until you see it in action.
Try this demo on a new worksheet:
In the first 3 cells (A1:C1) enter "whatever" random values. I used 1,2,3, like this:
In D1 enter formula: =OFFSET(D1,0,-2)
In E1 enter formula: =OFFSET(E1,0,-2)
Note that both formulas refer to themselves. With most functions, this would create a circular reference error, but in this case, it's not looking for a value in that cell, just a starting point.
After the cell reference, the next two values in the function are the number of rows, and then number of columns, to offset by. Negative numbers would count "up" for rows, or "left" for columns, and positive numbers count "down" and "right".
Technically, OFFSET can take 2 other values too (for 5 parameters in total) but in this case, you can ignore the last 2. (They are to specify height and width, irrelevant unless you get into working with array formulas.)
So, the function in D1 is offsetting from cell D1 (itself), by zero rows (so neither up nor down), and "2 columns to the left".
→ And now, if you drag D1 and E1 elsewhere (or insert columns/rows, etc) those 2 cells will automatically count the offset from the new location.
Just make sure you keep them as relative references (so D1, not $D$1) or else it will count from D1 regardless of where you move the formula.
...clear as mud? :-)
As an afterthought, here are a couple of other ways to visualize how OFFSET counts the cells to offset:
Use a lookup table (Insert > Table from the ribbon, or alternatively create an ordinary range and use cell references instead of the table syntax in the formula below).
Value
Abbreviation
Newbie
N
Promise
P
Waning promise
Wp
...
...
From this, the value in your Wk2-Wk3 column can be calculated with a VLOOKUP(), XLOOKUP() or INDEX(MATCH()). For example,
=VLOOKUP(C3,LookupTable[[Value]:[Abbreviation]],2,FALSE)&"-"&VLOOKUP(D3,LookupTable[[Value]:[Abbreviation]],2,FALSE)
where LookupTable is the name of the table (you can set this on the "Table Design" ribbon tab).
Well, I did it like this, using text functions:
IF(IFERROR(FIND(" ",C3,1),0),LEFT(C3,1)&LOWER(MID(C3,FIND(" ",C3,1)+1,1)),LEFT(C3,1))&"-"&IF(IFERROR(FIND(" ",D3,1),0),LEFT(D3,1)&LOWER(MID(D3,FIND(" ",D3,1)+1,1)),LEFT(D3,1))
And, if you add the new week data columns appropriately then all you need to do is drag right...
Related
So this needs a bit of detail:
n,X,X,X,n is in cells B5 to F5
I need to get the following output:
1n,3x,1n
for this particular row.
Now the n's and X's represent stitches in knitting with the "n" being the background color and the "x" being the front color.
There is an array of cells B5:F12 representing the rows and stitches, so each row will have a different arrangement of stitches or background color.
I need to avoid vba as this needs to be as stable as possible with the user being my Mum who is 90 years old :) and all she needs is a place to enter the name and the layout (which I have done) and a pattern list for each row (also sorted).
I have started to consider things like:
if(B5=C5,1&B5,"")
But given the n umber of combinations that becomes very long.
Any ideas? Cheers.
You could try:
Formula in H5:
=BYROW(B5:F12,LAMBDA(x,LET(z,REDUCE(VSTACK(TAKE(x,,1),1),DROP(x,,1),LAMBDA(a,b,IF(b=#TAKE(a,,-1),IF(COLUMNS(a)=1,VSTACK(b,TAKE(a,-1)+1),HSTACK(DROP(a,,-1),VSTACK(b,DROP(TAKE(a,,-1),1)+1))),HSTACK(a,VSTACK(b,1))))),TEXTJOIN(",",,DROP(z,1)&TAKE(z,1)))))
I'll see if I can golf the bytecount down a bit...
EDIT:
After a considerable amount of golfing (came down to 119), I came up with:
=BYROW(B5:F12,LAMBDA(x,MID(REDUCE("",x,LAMBDA(a,b,IF(RIGHT(a)=b,LEFT(a,LEN(a)-2)&1+LEFT(RIGHT(a,2)),a&",1")&b)),2,99)))
Though less dynamic than the 1st one, but possible due to the fact there are only <10 columns for each knitting pattern.
If your mother doesn't have the latest Excel (with LAMBDA etc), here is an alternative to #JvdV's answer which only uses LET,SEQUENCE and FILTER.
It only accepts a single row, so you'd need to fill the formula down.
=LET(p,LOWER(B5:F5),c,COLUMNS(p),s,SEQUENCE(,c),
a,IF(s=c,c,IF(INDEX(p,,s)<>INDEX(p,s+1),s,0)),
b,FILTER(a,a>0),t,SEQUENCE(,COLUMNS(b)),
n,IF(t=1,INDEX(b,,t),INDEX(b,,t)-INDEX(b,,t-1)),
TEXTJOIN(",",TRUE,n & INDEX(p,,b)))
I might add that it allows for adding more than one colour into the pattern ...
and with a bit of conditional formatting, the good lady can design her own multicolour patterns!
This is just a start of a solution, but in cell "B6" you can put the formula:
=(IF(B5=A5,A6+1,1))
This will generate following list:
B C D E F
5: n x x x n
6: 1 1 2 3 1
From there, you can try to get the Subtotals feature to work, based on the Max formula, ... (as I said, this is just a start).
If you are willing to spread the logic over multiple sheets, it's quite easy to come up with a way to do this. Consider a workbook with three sheets:
Pattern
EqualPrevCol, where each cell of Pattern is checked for equality against the previous column of the same row.
The formula for cell EqualPrevCol!D3 is:
=Pattern!D3=Pattern!C3
And finally PatternResult, where most of the logic resides:
Consider one row of EqualPrevCol:
At every FALSE column, we want to know how many columns until the next FALSE. To do this, we want to find the next exact MATCH for D3 in the rest of the row:
=MATCH(EqualPrevCol!D3, EqualPrevCol!E3:$H3, 0)
If no match is found, that means the rest of the row is all TRUE. In this situation, we want to return the length of the rest of the row plus this current cell.
=IFNA(MATCH(...), COLUMNS(D3:$H3))
And finally, we append this to the current character:
=IFNA(...) & Pattern!D3
Also, if the 7 row at this column is TRUE, we want to keep this blank:
=IF(EqualPrevCol!D3, "", IFNA(...) & ...)
The full formula of cell PatternResult!D3 is:
=IF(EqualPrevCol!D3, "", IFNA(MATCH(EqualPrevCol!D3, EqualPrevCol!E3:$H3, 0), COLUMNS(D3:$H3)) & Pattern!D3)
Finally, the pattern is condensed to the Pattern sheet. The Pattern!B3 cell contains:
=TEXTJOIN(", ", TRUE, PatternResult!D3:$H3)
To scale this up, you simply need to change all occurrences of $H in the formulas (this was a reference to the last column) and re-fill the cells on the latter two sheets.
I have a small data set of 2 columns and several rows (columns A and B)
I want to return each instance of codeblk 3 in a formula that is elsewhere in my sheet, (so a vlookup is out as it only shows the first instance) if it does not appear then a message to say its not there should come up.
I have the formula partially working but i cant see the reason why its not displaying the values.
My formula is as below:
This is an array
{=IF(ISERROR(INDEX($A$55:$B$70,SMALL(IF($B$55:$B$70=3,ROW($B$55:$B$70)),ROW(1:1))-1,1)),"No value's produced",INDEX($A$2:$C$7,SMALL(IF($B$55:$B$70=3,ROW($B$55:$B$70)),ROW(1:1))-1,1))}
The result that shows up is only "No values produced" but it should reflect statement B, C and D in 3 separate cells (when changing ROW(1:1), ROW(2:2) etc)
{=SMALL(IF($B$56:$B$69=4,ROW($B$56:$B$69)),ROW(1:1))} - This produces the result 68 which is the correct row.
Any ideas?
Thanks,
This is an array formula - Validate the formula with Ctrl+Shift+Enter while still in the formula bar
=IFERROR(INDEX($A$55:$B$70,SMALL(IF($B$55:$B$70=3,ROW($B$55:$B$70)-54),ROW(1:1)),1),"No value's produced")
The issue you are facing is that your index starts it's first row on $B$55, you need to offset the row numbers in the array to reflect this. For example, the INDEX contains 16 rows but if you had a match on the first row you are asking for the 55th row from that INDEX(), it just can't fulfil that.
EDIT
The offset was out of sync as your original formula included another -1 outside of the IF(), I also left an additional bracket in play (the formula above has now been edited)
The ROW() function will essentially translate $B$55:$B$70 into ROW(55:70) which will produce the array {55;56;57;58;59;60;61;62;63;64;65;66;67;68;69;70} so the offset is needed to translate those row numbers in to the position they represent in the indexed data of INDEX().
The other IF() statement then produces and array of {FALSE;2;3;4;FALSE etc.
You can see these results by highlighting parts of the formula in the formula bar and hitting F9 to calculate.
Currently. I have it set up when someone enters information in column E (Action Taken) then the whole row will turn yellow so it is easily seen that that item is being dealt with.
Another conditional format that I have set up is that once 5 working days pass on the date entered into column F (Date Actioned) then the row will turn red alerting the user to chase up the issue again.
Every week the user will contact more suppliers on the list regarding the listed product, and fill in their action taken and date actioned.
What I am looking to do is to add up all the suppliers contacted in one week (i.e. week 1 , 05.02.18 to 09.02.18) Which I know can be done by using the formula;
=COUNTIFS(F4:F18,">=5/2/18",F4:F18,"<=9/2/2018")
HOWEVER I only want to count each company once! So even though between the date 05.02.18 to 09.02.18, 8 actions were carried out, they only contacted 4 suppliers.
Is this possible? (FYI the screenshots attached are just a quick mock-up of the real document which contains thousands of products and more in-depth information).
If you don't mind adding a couple of working columns, the following is a possible solution:
Step 1
Set up two cells to contain the min and max dates for your range that you used in your formula =COUNTIFS(F4:F18,">=5/2/18",F4:F18,"<=9/2/2018"). This will allow you to up date things without having to recopy the formula each time. I used cells E1 and G1 for min and max date respectively.
Step 2
In a new column generate a list of supplier IDs that match your criteria. I arbitrarily chose column H. I placed the following formula in H4 and copied down:
=IF(AND(F4>=$E$1,F4<=$G$1),A4,"")
Step 3
In a new column generate a list of the count of the first time a supplier ID occurs from column H and do not count blank spaces. I arbitrarily chose column I. I placed the following formula in I4 and copied down:
=--(AND(COUNTIF($H$4:H4,H4)=1,H4<>""))
Step 4
Take the sum of the results from step 3. I arbitrarily chose to place the following formula in I19:
=SUM(I4:I18)
The "cop out" route, if you were only doing the calculation for 1 week would be to have a Pivot Table, with the Company as the Rows and the Date as the Columns. Then filter your columns, and you get a row per Company that week.
The Full route means looking into Array Formulae - you type one of these like normal, but instead of pressing [Enter], you press [Ctrl]+[Shift]+[Enter], and the equation will show in curly braces. (This is why they are sometimes called "CSE Formula" for "Control, Shift, Enter")
~Warning - This is a long walk through how I calculated the final formula. Hopefully it will make sense though!~
An array formula lets you iterate through every row in a range, calculate them seperately, and then add them all together at the end. This gives us the outside term - everything goes inside a =SUM(..). Press [Ctrl]+[Shift]+[Enter], and this will display as {=SUM(..)}
Now, what calculation do we want to be doing per row? Well, for every row where column F is within date, you want 1 per company within date. Well, another was to say "per" is "divided by" (hence the "Percentage" symbol being "0 / 0" or "%", and the lesser knows "Permille" symbol being "0 / 00" or "‰")
For the sake of reproducibility, I am going to assume that your Count is being done in Cell J2, and you have the Start/End of week dates in Cells H2 and I2
The formula for Row 4 would start out as =IF(AND(F4>=$H2,F4<=$I2), 1/???, 0), where ??? is how many times the Company on row 4 appears. Unfortunately, the AND function does not work with Array Formula. Fortunately there is an easy workaround - since boolean True/False can be treated as binary 1/0, the AND operator is the same as multiplication. Converting from boolean to binary is fastest by double-negation, so AND(F4>=H2,F4<=I2) can be rewritten as --(F4>=H2)*--(F4<=I2), which will work fine with array formula, giving us =IF(--(F4>=H2)*--(F4<=I2), 1/???, 0)
Now, before we work out what ??? should be, I'm going to skip down to row 18. There is method to this madness, I promise! Following the rules laid out above, the formula for row 18 is =IF(--(F18>=H2)*--(F18<=I2), 1/???, 0) - fairly simple? Now, if we want to run this for all rows from 4 to 18, we just need to join the ranges. So, you would get =IF(--(F4:F18>=H2)*--(F4:F18<=I2), 1/???, 0) Turning this into an Array Formula would get you an array of "1/???"s and "0"s, so add it together with SUM to get a single 'flat' number: =SUM(IF(--(F4:F18>=H2)*--(F4:F18<=I2), 1/???, 0))
Remember that "method" I mentioned earlier? Well, if we were looking for a straight Count of actions taken, ignoring the Unique Companies constraint, we could make ??? = 1, to get =SUM(IF(--(F4:F18>=$H2)*--(F4:F18<=$I2), 1, 0)) and hit [Ctrl]+[Shift]+[Enter] - which would give the same result as =COUNTIF(F4:F18, ">="&H2, F4:F18, "<="&I2) - and changing the 1 to D4:D18 would be a SUMIF on Products. (Incidentally, you can remove the IF statement from the CountIf, since the Condition returns 1 or 0 already)
So, last step! What is ??? Well, it's a Count of rows where Date Actioned is within our date bounds, and Company is the same as the current row. So, for Row 4 you have COUNTIFS(B4:B18, B4, F4:F18, ">="&H2, F4:F18, "<="&I2), and for Row 18 you get COUNTIFS(B4:B18, B18, F4:F18, ">="&H2, F4:F18, "<="&I2), making the final ??? become COUNTIFS(B4:B18, B4:B18, F4:F18, ">="&H2, F4:F18, "<="&I2) (The trick here is that the Range arguments are treated as all the cells at once, but the Condition arguments are treated as an array of each cell one-at-a-time)
Now, some people might argue that you need to put your 1/COUNTIFS(B4:B18, B4:B18, F4:F18, ">="&H2, F4:F18, "<="&I2) inside an IFERROR to be safe - however, this is in the TRUE part of an IF statement referrring to the same row, so we know that if the COUNTIF is being evaluated then there will always be at least 1 row that matches.
So, if we stick it all together, and add a load of $ symbols to lock the rows/columns in place (so that you can drag down column J to calculate for different weeks/dates in columns H and I) you get the (slightly unwieldy) formula as follows:
=SUM(IF(--($F$4:$F$18>=$H2)*--($F$4:$F$18<=$I2), 1/COUNTIFS($B$4:$B$18, $B$4:$B$18, $F$4:$F$18, ">=" & $H2, $F$4:$F$18, "<=" & $I2), 0))
And, one last time - Don't forget to press [Ctrl]+[Shift]+[Enter]
(The 2 Date cells, $H2 and $I2 are the only terms I have left 'unlocked', and even then only on the Row)
I'm working on a budget for a project with multiple phases. There is a possibility that not all phases will be worked on so I've added some lookups and SUMIF formulas so that I can get a summary of my included and excluded effort and dollar amounts. That all works fine. Now I'd like to hide my row of lookups (row 1), but still have a way of identifying which phases of the project are included and which are excluded. Obviously I could manually concatenate them together, but if the phases being included/excluded change then I need to remember to update those formulas (and it's not nearly as fun as doing all in a formula). Here's how my sheet looks:
The TEXTJOIN function seems like it should work (i.e. =TEXTJOIN(CHAR(10), TRUE, C2:N2)), but I can't wrap my head around how to make the range parameter dependent on my lookup row. I played around with INDEX using something like =TEXTJOIN(CHAR(10), TRUE, INDEX(A2:M2,,(A1:M1="Yes")*COLUMN(A1:M1))), but didn't have any luck. At the end of the day I want to have something like:
Phase 1 Phase 2 Phase 5
Please note that the above data should all appear in the same cell - using the line feed character, CHAR(10), as the delimiter in the TEXTJOIN function will make all of the phases appear on a new line within a single cell. I do not want to fill formulas through multiple cells. Thanks in advance for any help.
Please take a look at the picture. I had a similar issue in the past (and brought it to StackOverflow, at which point I was helped by #ScottCraner, original post here:
Doing an array formula lookup
Basically,
1) You set up the array you are looking through - in my case, its $A$2:$A$6, in your case it will be $B$2:$M$2.
2) You then nest the COUNTIFS function inside of a match function. The function does the following:
A) It checks whether the name of phase X has already shown up in column E when you are copy/pasting down
B) If it has, move on to the next one
C) If it hasn't, output
3) Of note: this is an array formula, so the formula itself is checking through every cell. So it looks at cell A2; if the cell in B2 is "Yes", and "Phase 1" isn't in range in column E, then put in phase 1. Because it is, its there. Then it looks at cell A3; if the cell in B3 is "Yes", the same thing for phase 2. Then it looks at cell A4; because B4 is "No", the *(B2:B6="Yes") will throw an error. and so on
4) Place error catching brackets around the function.
A less elegant way to go about this, which doesn't require array formulas would be to use the secondary column's "Yes/No" as an index. This assumes the value in the secondary column is redundant- and repeating the primary column's "Yes" or "No."
In C1: =CONCATENATE(B1,COUNTIFS($A$1:B1,B1)) -
Repeat relatively for E1, G1, etc.
Somewhere separate (the below formula assumes Row 1 of the same worksheet), you could then use: =IFERROR(INDEX($2:$2,MATCH(CONCATENATE("Yes",ROW()),$1:$1,0)-1),"")
To look up "Yes1", returning "Phase 1" and autofill downward. IFERROR is used here to return blank when you run out of "Yes" results.
I'm trying to use this formula:
=SUMPRODUCT(VLOOKUP(B$4778,$D$4:$DC$4623,{4,5},0))
It works fine but I'd like to try to use a variable for the {4,5} portion of the formula (columns in the array to be summed) as the formula needs to change based on sheet inputs before this formula.
I have cells on the sheet that are to be used to set the initial and final columns to be searched (likely 10 columns, but the 10 columns would have to be selected from 90 some columns available).The columns are populations related to each age. So, if I need population of those aged 10 through 15, I'd need to sum up 5 columns. If 20-25, need to sum up 5 different columns.
I tried to use the Columns function but it didn't seem to work for me.
The columns are selected by users entering in cells the upper and lower limits of the search range and then I convert those values to the corresponding numerical column value.
So if they select 5 as lower and 10 as upper limit, I know I have to add 7 to get the correct data column on the data page (column 12) and likewise for upper (column 17).
The entire possible area to search is $D$4:$DC$4623. So, in the formula, if I wrote it out long way it would be:
=SUMPRODUCT(VLOOKUP(B$4778,$D$4:$DC$4623,{12,13,14,15,16,17},0))
I'd prefer to write it out using variables, something like this:
=SUMPRODUCT(VLOOKUP(B$4778,$D$4:$DC$4623,{L:U},0))
Where variable L would be 12 and variable U would be 17.
Can anyone suggest a way to write the formula?
use this array formula:
=SUM(VLOOKUP(B$4778,$D$4:$DC$4623,ROW(INDIRECT(D5 & ":" & E5)),FALSE))
Where D5 is the Lower and E5 would be the upper.
Being an array formula it must be confirmed with Ctrl-Shift-Enter instead of Enter when exiting edit mode. If done correctly then excel will put {} around the formula.
Or better yet use this non array formula:
=SUM(INDEX($D$4:$DC$4623,MATCH(B$4778,$D$4:$D$4623,0),D5):INDEX($D$4:$DC$4623,MATCH(B$4778,$D$4:$D$4623,0),E5))