To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=76142


User pperson changed the following:

                What    |Old value                 |New value
================================================================================
               Component|Spreadsheet               |documentation
--------------------------------------------------------------------------------
              Issue type|DEFECT                    |ENHANCEMENT
--------------------------------------------------------------------------------
              QA contact|[EMAIL PROTECTED]                 |[EMAIL PROTECTED]
--------------------------------------------------------------------------------
                 Summary|LINEST Example Fails      |LINEST Help Info Sucks
--------------------------------------------------------------------------------
            Subcomponent|code                      |Online help
--------------------------------------------------------------------------------
        Target milestone|---                       |not determined
--------------------------------------------------------------------------------




------- Additional comments from [EMAIL PROTECTED] Sat Apr  7 17:45:53 +0000 
2007 -------
Well, the good news is I got it to work. This is now an enhancement request for 
the Help File. Huh, doesn't look like that is a Component or Subcomponent. Oh, 
well.

The less good news is that the Help File sucks, at least for the LINEST entry. 
Here are the details:

a) The statement 

Linear_Type (optional). If the line goes through the zero point, then set 
Linear_Type = 0.

should read:

Linear_Type (optional). To force the line to go through the origin, set 
Linear_Type = 0. Any other value (or no value) produces a non-zero constant 
(unless the line actually goes through the origin).

b) The example shows a blank column D. It is not necessary to have such a 
column. The actual restrictions, if any, on output array placement should be 
clearly given, either here or at the section introduction (which, by the way, 
reads like a man page, ie, is completely unintelligible unless you already know 
what it is saying).

c) The example shows two X columns, which is unnecessarily confusing. At a 
minimum, an example with one X columns should be added.

d) The output array should be shaded so that there is no doubt what it is. If 
your help system cannot do this, you need a competently-written help system 
that can.

e) The description of calculated values needs to be completely redone. Here is 
an example:

1) X column order: X columns are read from left to right within the block. This 
establishes the normal order.
2) The first row of the output array will have n+1 columns, where n is the 
number of X columns. The n+1st column contains the constant. Moving to the 
left, the slopes for each X column is found in reverse order: the first X 
column's slope is adjacent to and to the right of the constant, the second X 
column's slope is adjacent to and to the right of the first X column's slope, 
and so on.
3) The second row of the output array will have n+1 columns, as above. Each 
cell will contain the standard error of the corresponding cell in the first row.
4) The third row will have n+1 columns, but the last n-1 will have #NA because 
the programmer's are too stupid to produce a blank cell where no output is 
needed. The first cell has the RSQ and the second the F value from the variance 
analysis.
5) The fourth row is similar to the third row, except that the first cell has 
the standard error of the regression calculated for the Y value and the second 
cell has the degrees of freedom from the variance analysis.
6) The fifth row is similar to the third row, except that the first cell has 
the sum of the squared deviation of the estimated Y values from their linear 
mean and the second cell has the sum of the squared deviation of the estimated 
Y value from the given Y values.

I agree with the original author, however, that anyone who doesn't understand 
what the terms mean needs to study statistics before attempting to use the 
information. Except, perhaps, that "RSQ" might be replaced by "r-squared", 
unless, of course, "RSQ" is a standard mathematical abbreviation for this item.


---------------------------------------------------------------------
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to