View Single Post
  #4   Report Post  
KiwiSteve
 
Posts: n/a
Default

Thanks to both Dave and David for the two replies. Calculation is set
to automatic and I tried F9 a few times just to be safe. You can never
be sure when helping someone else after the fact, but the user says she
created two new adjacent columns using the insert command on the column
header and then dragged the existing formula and cell values from the
adjacent left column. At that point the new column had value errors.
After tracing the origin of the value error to one particular cell in
each column , I tried Format Cell Number format on all feeder cells
with no change. I looked at all of the Tools-Options tabs and their
settings and saw nothing that I recognised as affecting the results. At
the suggestion of a colleague I even copied the working and non working
cell formula to the clipboard and pasted them into Notepad. Apart from
expected column reference differences there were no differences to the
formula. I then eventually found that doing any of the following to the
non working cell fixed the problem. 1) Editing the cell in the edit bar
(with no change) 2) Using David's replace equals with equals edit 3)
Dragging the left hand working cell over the non working cell (which is
what the user says she did in the first place to fill the cell. Using
the format painter did not fix the problem. It really acts as if Excel
has internally lost track of what these non working cells are all about
and then resets itself once the cell is edited. Now that we know Davids
work around, its no big deal, but I would have liked to know the
conditions that trigger this in an attempt to avoid it in the future.

As requested the formula is =(CEILING(SUM(H35:H43),1)) where the range
has cells that have calculated results that are all working ok.

As an aside I have seen many times a similiar problem with data we
import into XL from external sources. Even though a cell contains a
numeric value and is formatted as numeric, XL will complain that the
cell has an error and wont calculate mathmatical results correctly
until we do the 'no change' edit. The only menu command that allows us
to fix the problem in mass is the Data-Text to columns command. The
other clumsy solution I have is a macro that loops through every
populated cell and does a non destructive edit on the value. Again XL
appears to be acting as if it has lost track of what is going on until
after the edit. The text to columns solution is a pain as I only know
how to apply it one column at a time and my macro is real slow (but
faster than doing it by hand). Once again, if I understood the failure
mode, maybe I could avoid it.