Unexpected changes in % complete column for summary

[Angela]

I have a large schedule in Proj.98, which has many
summaries and sub-tasks. One of my second level summaries
changed from 20% to 0%, and I can't figure out why or how
to get it back. I did the calculations, and there should
be a % there, but it shows 0%.

Thanks,
Angela

 

[John]

Angela,
Let me ask the obvious question first. Is Calculation set to "automatic?"

When you say you did the calculations, what do you mean? How did you
calculate the summary Percent Complete?

John

 

[Angela]

Yes, it is on automatic. I used the formula that you
gave me several weeks ago. The sub-tasks and summaries
within this summary level have completions and partial
completions (90%, 40%, 30%, etc.), so there should be
something here.

 

[John]

Angela,
Well if I gave you the formula, you know it's gotta be right, right?
Yup, I sounds like there should be a value on the summary line but I
don't have a ready answer. Would it be possible to send me either the
whole file or just the summary and subtasks where the phenomenon occurs?

John

 

[Angela]

John,
Due to proprietary info. I can't send the file to you,
but could it have anything to do with using different
versions of Project on the same file? Here's the deal. If
I open the file up in Proj. 98 the % on that line is 0%,
but if I open it up in Proj. 2002 it shows the correct
19%. I have to explain the to my supervisors, in order
for them to have confidence in Projects obility to give
the correct percentage.

 

[JackD]

I have a macro on my website which wipes names and other stuff clean so you
can send the file to someone else.
It preserves the heirarchy so you can figure out what is going on.
You could try downloading and running it.
http://masamiki.com/project/macros.htm
Look for the scrub macro.

-Jack

 

[John]

Angela,
Just for you reference, I can send you a macro that will remove all
proprietary information from a file. It basically makes everything
generic (i.e. task and resource names) and changes all cost values to
meaningless values (i.e $1/hr or $0). If interested, I send the macro to
you. BTW, Jack Dahlgren has a similar macro ("scrub") on his website at:
http://masamiki.com/project/macros.htm
It isn't quite as aggressive as mine but you could try it to get the
gist of what it does.

Off hand, I can't think of any reason why Project 98 would give a
different result than Project 2002 for %Complete on a Summary line. As
far as I know that calculation has been the same for ages (i.e. Project
versions). I can understand your supervisor's concern about data
validity when you get the type of results you get, however keep in mind
that %Complete at a Summary line is not always a good representation of
what is happening. For example, look what happens to %Complete on a
Summary line that has only milestones as subtasks.

If you don't want to try the macro(s) I suggest, can you replicate the
behavior in a small file that doesn't have sensitive info?

John

 

[JackD]

Angela,

After reading more of this thread I realized your problem.

Project 98 DOES NOT support formulas in custom fields.

This was a new feature in project 2000.

If you want to do the same calculations, you can still do them using VBA,
but you will have to run a macro each time you want it to recalculate.

-Jack

 

[John]

Jack,
Angela's "calculation" was a manual one. A few weeks ago she posted a
query asking how %Complete is calculated for a Summary line. I
corresponded with her both on and off this newsgroup and gave her the
"formula" for calculation %Complete for a Summary line. Apparently, she
has a file that for Project 98, a 0% value is showing on a Summary line
that has non-zero %Complete for its subtasks, but the same file shows
the correct "calculated" Summary line value when run on Project 2000.
I'm anxious to understand what is going on.

John

 

[Angela]

My schedule is currently contains 13,727 task items, and
the file is 31,611KB. Will your macro or Jack's work, or
will it crash? I did find that I could open the file in
2002 where the correct % shows up. Then save the file
from there back as a 98 file, but it still puzzles me
what happened to begin with.

Thanks,
Angela

 

[JackD]

My macro should not crash just because the project is big. It might take a
bit longer though.
However, my macro IRREVERSIBLY strips information from the project. For this
reason use it only on a COPY of your file.

I sometimes question why people need a schedule with 13,000 tasks.
One wonders if a manager can actually comprehend and manage such detail or
whether it would be better to have a heirarchy of schedules - each of them
with fewer tasks so that one can focus on what the real issue is.

-Jack

 

[John]

Angela,
No crash with my macro either but I have to agree with Jack, if you (or
your management) is trying to manage a single Project file with 13K+
tasks, you've already lost the war. The most I have worked with is a 10K
file and that was a master file made up of 70+ individual subprojects.
Working with that large a file wasn't easy and I have a ton of
experience.

Nonetheless, I gotta believe that the problem you are seeing can be
replicated with a very small portion of the huge file. Did you try
copying just the Summary line that shows the problem along with its
subtasks and paste it into a new file?

John

 

[Angela]

I agree, but we are not looking to manage each project
via this method. Once they decide the best way
to "distribute the wealth", we will break it down for
each area to manage their own sections. Right now the
main function is the "Big picture" resourcing and timing
efforts.

 

[JackD]

When each task is less than 1/100 of one percent of the project, it takes a
lot of work to do the big picture resourcing. How can you know you are
correct unless you look at each of those 10K tasks.
I venture that you would lose less than a percent or so accuracy (and may
gain some) if you reduce the number of tasks by a factor of 10.
Especially if you are working at the "big picture" level. When you break it
down have each area build up any necessary detail.

-Jack