Why input range contains non-numeric data?

  • Thread starter Thread starter Guest
  • Start date Start date
G

Guest

When I try to do a regression analysis on my data, a box pops up saying there
is non-numerica data for the input. However, I cant find any within my data
sheet.
 
When I try to do a regression analysis on my data, a box pops up saying there
is non-numerica data for the input. However, I cant find any within my data
sheet.
Click to expand...

Are some of the cells formatted as text? Or entered with a leading
apostrophe (which amounts to the same thing)?

Mark Lincoln
 
Mark Lincoln said:
Are some of the cells formatted as text? Or entered with a leading
apostrophe (which amounts to the same thing)?

Mark Lincoln

I selected all of the data and then formatted the cells as numbered. And as far as apostrophes go I cant find any. But the problem still persists.
Click to expand...
 
Reformatting a cell does not change the value in the cell, it merely changes
the display of that value. In particular applying a numeric format will not
change text into a number. Copy an empty cell, select the input range, and
Edit|Paste Special|Add to coerce the values into numbers.

To identify which cell(s) is(are) text,
Format|Cells|Alignment|Horizontal|General. Text will be left justified while
numbers will be right justified. Alternately COUNT() only counts numbers.

Jerry
 
If all your data look like numbers and still you have a problem. Excel need
to calculate with them to understand that its numbers. Try to write 1 into a
cell and Copy-Paste Special and multiply all the data. Then, it should work.
Alternatively you can hit F2 in every single cell. Same idea.
Gunnar
 
Numbers formatted as text have to be converted back to numbers.
Create a range of cells in the same shape as the problem data cells,
each of which contains a 1. Then Copy this range, select the top-left
cell of the problem data range, select Paste Special from the Edit
menu, click on the Multiply button in the Operations section of the
Paste Special dialog, then click OK. Your numbers will now be treated
as numbers by Excel.

Mark Lincoln
 
Ok, so I copied and pasted special with a blank box and a bunch of random
numbers in my columns went right justified while most stayed left. But that
didn't fix anything. Then I tried all of the other suggestions and still
nothing it working. I still have random numbers sitting on the right and
cant do any data analysis. This is getting frustrating.



Nickman
 
I think you might have space infront or behind some of your data. Use F2 in
some of the cells to make sure that there not are any "odd characters"
unvisible.
You can check this easy also by using =Left(b1,6) where 6 is numbers of
characters
you want to extract.
This formatting problems is very irritating in Excel. Numbers are not always
calculatable(if thats a word) even if its seems like it.
Good luck
G.
 
Hi,
I am new to this forum and new to the Data Analysis Tool in Excel 2003.
I have a file consisting of 11,294 rows of data in 6 columns. 5 of the columns contain text like full name and one column contains SSN that I have converted to the Number format. I am trying to get 114 periodic samples out of this population.
When I use the Data Analysis Tool, I put in the input range $a$1:$f$11294, I select periodic and put in 100 for the interval of sampling and I want to put it into my Audit Sample worksheet.
Everytime, I do this I get the message: "Sampling - input range contains non-numeric data".

I tried to do the Copy of one and Paste special but I still get the
same message.
Is this not the correct tool to easily pull out every 100
sample from my file or am I using it incorrectly? Please any help would greatly be appreciated!
 
Also, check to see if any columns or rows are hidden in your ranges. I had forgotten two rows of non-numeric values were hidden on my worksheet.
 
I had the same problem and tried to use those solutions and it didn't work...

But THEN I noticed that some rows were hidden and didn't show up on the filters. So I simply eliminated them and then it al worked out.

One way to try to locate those rows is to "attempt" calculations of porcions of the data table. For example, if you can calculate with rows 1 to 100 and with rows 140 to 180, but cant calculate in between 100 to 140, the error is in that span. This way you will spend less time looking for the error.

I dont know however if this will work with all errors, I just happened to notiice those rows.:thumb:
 
Back
Top