Update: Through some additional testing I discovered:
1) 255 Characters does seem to be the breaking point (character limit). Setting the filter with an array with a character length of 245 worked fine -- I was able to save and reopen without any errors. I added another criteria to the array to make the length 262, saved the file, and then got the same error.
2) The sheet in the removed records message refers to the sheet index, not the sheet name, and it does indeed reference the sheet with the autofiltering. End Update
My Issue -- I've written code to set a dataset's AutoFilter based on selected items in several slicers. Sometimes when I open up the file I get the error (paraphrased): Excel found unreadable content in the workbook. Do you want to repair the file? Then a dialog pops up and says Removed Records: Sorting from /xl/worksheets/sheet2.xml part.
The code works as designed; the dataset reflects whatever is selected in the slicers (even many selections).
I set the array (a string array) as follows and then use the array to set the criteria:
If sCache.Name = "Slicer_Test" Then
For Each sItem In ActiveWorkbook.SlicerCaches(sCache.Name).SlicerItems
If sItem.Selected = True Then
ReDim Preserve sArr(0 To sCount)
sArr(sCount) = sItem.Name
sCount = sCount + 1
End If
Next sItem
filterRng.AutoFilter Field:=9, Criteria1:=sArr, Operator:=xlFilterValues
ReDim sArr(0 To 0)
End If
I replicate the above code for each slicer.
Where I think the problem stems from is that the three largest slicers contain 27, 120, and 322 items, respectively. So as you can imagine, when all the items in the largest slicer are selected, the array's string length is over 5K characters long... like I mentioned above, the code works as designed. I found this thread, which mentions a character maximum?
I've tried removing the filters before saving/closing the workbook, but that doesn't always work, and this file will be used by many other people. So I'm wondering if 1) anyone has a suggestion for a way to workaround this error, or 2) if there might be a way to accomplish the filtering without using a terribly-long array...
Any thoughts on this will be much appreciated!
A co-worker of mine helped me resolve the issue.
Apparently when using this syntax:
Criteria1:=sArr
Excel reads the array as one long string instead of looking at it as an array that contains many string elements.
The fix is to use the Array() function like so:
Criteria1:=Array(sArr)
This seems to prevent Excel from corrupting.
Sorting before autofilter will help you to perform autofilter function faster and better
Related
I am working on a problem in Excel, in which I have to compare data in two separate workbooks, and look for matching pieces of information over several columns - as the time stamp in row A isn't an exact match in both workbooks I need to rely on the data points from the other columns out to column G - and format the rows in which a match is found in a certain colour.
I could do this manually, but due to the amount of rows numbering into 5 figures, I think VBA seems the best way to do this. Having only basic programming knowledge, I am struggling to see the correct way to do this with VBA.
Here is what I have tried so far, and I will also include the error shown when I run the code.
I am aware I said that more than row A is needed to make the comparison, but I am not sure on how to apply the rules to a matrix over columns and rows.
Sub vbax_53997_Compare_Two_Ranges()
Dim i As Long
Dim wb1ws1, wb2ws2
Dim blnSame As Boolean
wb1ws1 = Workbooks("Copy_of_data.xlsm").Worksheets("Worksheet").Range("A1:A63").Value
wb2ws2 = Workbooks("Copy_of_data_to_be_compared.xlsm").Worksheets("Archive").Range("A1:A23067").Value
For i = LBound(wb1ws1) To UBound(wb1ws1)
If wb1ws1(i, 1) = wb2ws2(i, 1) Then
blnSame = True
End If
Next i
If blnSame = True Then
Sheets(sheetName).Cells(lRow, "A").Interior.ColorIndex = 3 'Set Color to Red'
End If
End Sub
This is the error when I run the code;
Run-time error '9': Subscript out of range.
I have no doubt its an easy fix and I could maybe do a nested vlookup/if statement, but taking longer in the short term to find a solution will probably be a benefit to automate the process in the long term.
Any help would be much appreciated as I have now run out of ideas.
So I have a large suite of code that creates an archive of data in sheets used by employees. Part of what makes this functional is named ranges on each sheet of usable data. In order for the data integrity to remain, I need to copy the named range objects from the archive sheet to its copy. The named ranges are built programatically and function as expected on the sheets. The problem I'm having is when I go to archive the sheet. Here is the code I'm using to handle the named range object:
For Each n In OldSht.Names
NamedRangeRefersTo = n.RefersTo
NamedRange = n.Name
TrimmedName = Right(n.Name, Len(n.Name) - InStr(1, n.Name, "!", vbTextCompare))
OldSht.Names(n.Name).Delete
OldSht.Names.Add Name:=ArchiveNamedRange, RefersTo:=NamedRangeRefersTo
Next n
The strings that grab data from n are used to add the same name object to the new sheet.
The problem I'm having is when a named range is referencing too large of a range when it hits the line Oldsht.Names.Add, it returns error 1004. I figured out it was the size of the referenced range by messing around with it. I haven't found the exact triggering cause, but this code works as-is when I use it on most of the named ranges. On large data sets with a joined data type that results in a very large named range (it would take a long time to explain how the ranges are built in text. It's a group of 8 sub functions with over 2000 lines of code), this results in the 1004 error.
What I'm confused by is why I can build the named range, use the named range, and copy the named range without issues (if I comment out the offending line, it executes perfectly but I lose data integrity). But when I take the referenced range into a code value, delete the old name reference, then add a new name(with a different name) and assign it the same refersto value of the old name, it can have this problem. I don't understand how it would be different doing this rather than just copying/renaming the name object. Unfortunately, I haven't found a workaround as of yet, nor have I found a clear cause of this error other than the fact that when I remove data or use smaller sets of data in test scenarios, I never have the problem. Does anyone have any ideas of what I can do? Does anyone have any ideas how a named range could be referring to a small enough range that it can be created, but using its refersto value to create a new named range could cause errors only when that is referencing a large range?
I wish I could provide some more concrete examples but unfortunately it would be very difficult to scrub enough sensitive info to provide the full code that would be necessary to reproduce my exact scenarios. Any ideas would be much appreciated.
As requested here's where ArchiveNamedRange gets set:
If Len(OldSht.Name) > 21 Then
ArchiveShtName = Left(OldSht.Name, 21) & DatePart("m", Date) & DatePart("d", Date) & DatePart("yyyy", Date)
Else
ArchiveShtName = OldSht.Name & DatePart("m", Date) & DatePart("d", Date) & DatePart("yyyy", Date)
End If
ArchiveNamedRange = ArchiveShtName & NameObjectName & "Test"
NameObjectName is just the name of the type of object and is passed in from another function. I'm not having an issue with the name just fyi. In the most extreme example the ArchiveNamedRange value at debug run time is = "OutageSystemProcedureMMDDYYYYSecurityRedactionTest" so the name might reach 50 and if things get crazier it might run upwards of 60 characters but it won't ever go beyond that or come anywhere near the 255 character limit. Ultimately, I haven't seen ArchiveNamedRange have an invalid value. It's just a string and it always has a value.
Edit-
Through my troubleshooting I've found that my code works when NamedRangeRefersTo has a length of 2075, but does not work when it has a length of 2091. So somewhere between 2075 characters and 2091 characters is a breaking point for assigning a string to RefersTo: in a named range.
So let's just assume there is a character limit for some reason of 2080 (or whatever it actually is between 2075 and 2091). When I initially find and create these named ranges, they are being given a range object. When I am copying the ranges I am copying as a string. Somehow when I pass a ranged object into RefersTo: it accepts characters beyond 2080 but when I pass in a string it does not. Given that this is my only breaking point of a large suite of code I'd rather find a workaround for this than have to re-factor the entire concept of my archive system. If I use a range object for copying the named ranges, their references follow the old Sheet. That means that when I copy the name over it can be "CriticalSystemsTest1" and referto: "CriticalSystemsTest1!$A$2,..." but once I copy that over and rename the archive worksheet (now CriticalSystems562015) the references adjust to be "CriticalSystems562015Test1!$A$2,..."
So I had to copy as a string to avoid that problem (it breaks data on the new sheet). All I really need is a creative way to overcome this character limit issue on my string. Rebuilding the named range from scratch on the new sheet is also not going to work. So I guess if anyone has ideas for how to work around this string size issue or a way of trimming the string while maintaining functionality of the named range, that would be amazing.
Each of these names has a worksheet level scope, so maybe if there's a way of using just the cell address($A$2) in RefersTo: so it doesn't also contain the worksheet reference (SheetName!), that would be a potential solution but I haven't figured out if that's even possible.
The reason the range definitions as strings are so long is that there are many areas within them. So one workaround would be to build up a new Range object area by area. You can use the string address of each area without running into any limits as each area only has a short reference. Using Range.Address gets the cell reference without the sheet reference, so you can create a new Range on a different sheet but with the same cells. Then use Union() to join all the areas and create the new name using the newly built Range instead of a string:
Dim i As Long, oldRange As Range, newRange As Range
Set oldRange = n.RefersToRange
Set newRange = oldSht.Range(oldRange.Areas(1).Address(External:=False))
For i = 2 To oldRange.Areas.Count
Set newRange = Union(newRange, oldSht.Range(oldRange.Areas(i).Address(External:=False)))
Next i
oldSht.Names.Add Name:="ArchiveNamedRange", RefersTo:=newRange
A couple of notes:
For ranges with many areas this is slow. If you can reliably tie down the threshold where you have problems, it would probably be worth testing for this first and only using this workaround where it was needed.
When testing I also ran into problems with using Worksheet.Range("some very long string range reference"), so this limitation isn't confined to named ranges.
I've been working all week to prepare a VBA application, which I'll be using in a meeting today. Unfortunately the code that has been running all week last week without a hitch, has decided to break over the weekend.
I constantly get Object variable or With block variable not set Run-time error '91' from this statement:
With Sheet5
Set adjrng = .Range(.Cells(.Range("G43:G60").Find(.Range("H39").Value).Row, 10), .Cells(.Range("G43:G60").Find(.Range("H39").Value).Row, 21))
End With
idea is to set a range in the row of the Range G43:G60 where the Value of H39 matches from Column 10 to Column 21.
Anybody spot the issue? My brainz are to nervous and sleepy this morning...
Thanks a bunch
Ben
EDIT:
After playing a bit with find and replace, the issue seems to be that excel has not yet properly calculated the "lookin" and "lookup" Ranges G43:G60 and H39. A simple recalculation didn't make excel rediscover the contents but when I used one of my input toggles to display a different value in those cells, and the went back to the original it did manage to find it.
Maybe using find for this is bad style, the find formula has these kind of hicups usually or any other comments on this? For now everything works fine again, but I'm afraid of running into these issues again. Any tips would thus still be appreciated.
EDIT: (from comment below)
We have a dynamic range (G43:J60) where unique identifiers are listed in column G and data is to the right. if something is changed in the data part of the range AND the lines uniqued identifier in column G matches the one in cell H39 a sub() is triggered via worksheet_ change intersect(target, adjrng) Defining that adjrng is the part that throws errors when find returns null.
I believe you are simply trying to set a range hoping that there will be two matches to the value in H39 within the G43:G60 range. While I avoid on Error Resume Next (never could adjust to the logic of breaking something in the hope to accomplish something), I always check that the values will be there when I look for them.
Dim rwUNIQ as long
Set adjrng = nothing
With Sheet5
if cbool(application.countif(.Range("G43:G60"), .Range("H39").Value)) then
rwUNIQ = application.match(.Range("H39").Value, .Range("G43:G60"), 0)
Set adjrng = .Cells(42 + rwUNIQ, 10).resize(1, 11)
end if
if not adjrng is nothing then
'do something with adjrng
end if
Set adjrng = nothing
End With
That checks to make sure that there are at least two H39 values in G43:G60 before proceeding. There is no further error control because we've counted at least two of them. You might want to compensate with an Else for when there isn't. If a single H39 value was found, you also might want to select a single row.
Remember that the .Find uses many parameters that were retained from the last time Find was used, whether in VBA or with a user on the worksheet. You have a real lack of parameters that specify the options that Find should use to proceed. e.g. xlPart or xlWhole, After:=what?, look in formulas or values, etc.
EDIT:* Modified the code to look for a single instance of the value in H39 and .Resize to expand the width (as per OP's comments).
I have a table of data with the top row being filters, I have a loop that changes which filter needs to be used inside the loop is the variable filterColumn that is being assigned a new value every time the loop runs through.
when i try to use filterColumn to determine which filter will be 'switched on' i get an error
Autofilter method of Range Class Failed
ActiveSheet.Range("$U$83:$CV$1217").AutoFilter Field:=filterColumn, Criteria1:="<>"
What is the correct syntax in order to use a variable to determine which field the filter is in?
Problem Solved I found the solution. I was referencing the filters columns position in terms of the whole worksheet when in fact I should have been referencing what number it was in the group of filters. For example the filter I wanted to change was in 'CF' which is the 84th column but my the filter I wanted to change is the 64th in the group.
Dim filterColumn As Integer
filterColumn = 2
ActiveSheet.Range("$U$83:$CV$1217").AutoFilter Field:=filterColumn, _
Criteria1:="<>"
EDIT: I tried #HeadofCatering's solution and initially it failed. However I filled in values in the referenced columns and it worked (my solution also failed under reverse conditions - make the column headers blank and it fails).
However this doesn't quite mesh with what I've (and probably you've) seen - you can definitely add filters to columns with blank headers. However one thing was consistent in the failures I saw - the filterColumn referenced a column that was outside of Application.UsedRange. You may want to try verifying that the column you are referencing is actually within Application.UsedRange (easy way: run Application.UsedRange.Select in the Immediate Window and see if your column is selected). Since you are referencing a decent amount of columns, it is possible that there are no values past a certain point (including column headers), and when you specify the column to filter, you are actually specifying something outside of your UsedRange.
An interesting (this is new to me as well) thing to test is taking a blank sheet, filling in values in cells A1 and B1, selecting columns A:G and manually adding AutoFilters - this will only add filters to columns A and B (a related situation can be found if you try to add filters to a completely blank sheet).
Sorry for the babble - chances are this isn't even your problem :)
Old solution (doesn't work when conditions described above are used)
I may be overkilling it, but try setting the sheet values as well (note I used a sample range here):
Sub SOTest()
Dim ws As Worksheet
Dim filterColumn As Integer
' Set the sheet object and declare your variable
Set ws = ActiveSheet
filterColumn = 2
' Now try the filter
ws.Range("$A$1:$E$10").AutoFilter Field:=filterColumn, Criteria1:="<>"
End Sub
In my excel VBA code, I need to move some data from a range to another sheet.
As of now, I'm iterating through the range and copying the values like this:
For offset = 0 To 101
ActiveWorkbook.Sheets(Sheet).Range("C3").offset(offset, 0).Value = ActiveSheet.Range("D4").offset(offset, 0).Value
Next offset
However, it takes almost a minute to iterate and copy the values for the 100 cells.
Would I be better off using Copy-Paste programatically, or is there a way to copy for the entire range at once? Something like:
ActiveWorkbook.Sheets(Sheet).Range("C3:C102").Value = ActiveSheet.Range("D4:D104").Value
You can read the entire range at once into a Variant array, and then write it back to another range. This is also quick, flickerless, and has the added bonus that you can code some operations on the data if you are so inclined.
Dim varDummy As Variant
varDummy = ActiveSheet.Range("D4:D104")
' Can insert code to do stuff with varDummy here
Workbook.Sheets(Sheet).Range("C3:C103") = varDummy
This I learned the hard way: Avoid Copy/Paste if at all possible! Copy and Paste use the clipboard. Other programs may read from / write to the clipboard while your code is running, which will cause wild, unpredictable results.
Also, it's generally a good idea to minimize the number of interactions between VBA and Excel, because they are slow. Having such interactions in a loop is multiply slow.
So, silly me did not try before posting here. Apparently, I can move data for an entire range this way:
Workbook.Sheets(Sheet).Range("C3:C102").Value = ActiveSheet.Range("D4:D104").Value
Its as fast as copy-paste without the switching of sheets.
Iterating through the range using a for loop takes about 45s for 100 cells, while the above two options are instant.
You can speed up code and stop flickering with:
Application.ScreenUpdating = False
'YOUR CODE
Application.ScreenUpdating = True
More: http://www.ozgrid.com/VBA/excel-macro-screen-flicker.htm
Columns("A:Z").Select
Selection.Copy
Sheets("Sheet2").Select
Range("A1").Select
ActiveSheet.Paste
That will copy columns A to Z from Sheet 1 to Sheet 2. This was generated by recording the macro. You can also apply it to ranges with something like this:
Range("D4:G14").Select
Selection.Copy
Sheets("Sheet2").Select
Range("D4").Select
ActiveSheet.Paste
Is this something like what you're after?
If you need anything specific and you can do it manually (e.g. copy and paste), record the macro to get the VBA code for it.
Copy and pasting has a decent amount of overhead in VBA, as does dealing with ranges like that. Its been a while since I have done VBA but if I recall correctly the fastest way to do something like this is to write the values you want into an array and then use the Resize function. So something like this:
Option Base 0
Dim firstrow as integer
Dim lastrow as integer
Dim valuesArray() as Long
Dim i as integer
//Set firstrow and lastrow however you deem appropriate
...
//Subtracing first row from last row gets you the needed size of the 0 based array
ReDim valuesArray(lastrow-firstrow)
for int i = 0 to (lastrow-firstrow)
valuesArray(i)=Cells(i+firstrow, COLUMNNUMBER).value
next i
Of course replace COLUMNNUMBER with whatever column it is you are iterating over. This should fill your array with your desired values. Then pick your destination cell and use Resize to put the values in. So if your destination cell is D4:
Range("D4").Resize(UBound(valuesArray)+1, 0).value = valuesArray
That write all the values in the array starting at D4 and going down to as many cells are in the array. Slightly more complicated but if you are going for speed I don't think I have ever come up with anything faster. Also I did this off the top of my head so please test and make sure that you don't cut off a cell here and there.
That OZGrid page has very useful info - http://www.ozgrid.com/VBA/SpeedingUpVBACode.htm
In my case, I need the formatting to be copied as well so I have been using this:
Sheet1.Range("A1:A200").Copy Destination:=Sheet2.Range("B1")
but was still having very slow execution - to the point of locking up the application - I finally found the problem - at some point in the past a number of empty text boxes got into my page - and while they were copied each time my code ran they were not erased by my code to clear the working area. The result was something like 4,500 empty text boxes - each of which was copy and pasted by even the code above.
If you use Edit - Go To... - Click on Special - then choose Objects - and you don't see anything that is good - if you see a bunch of objects that you were not aware of on your page that is not good.