PH conventions
Karen Barrett
kayandjade at aol.com
Fri Mar 6 09:38:49 CST 2015
There is in fact a convention being held in CA soon by the new owner of PH, Unicom Global. http://unicomsi.com/files/3014/2427/4609/UNICOM_Global_CAB_and_UGMs_at_PickFair.2015.03.27.pdf
Karen Barrett
-----Original Message-----
From: powerh-l-request <powerh-l-request at lists.sowder.com>
Today's Topics:
1. RE:Hierarchical vs. Parallel linkage
(Matt
Ohmes)
----------------------------------------------------------------------
Message:
1
Date: Fri, 6 Mar 2015 08:52:06 -0600
From: Matt Ohmes
<mohmes at us.ibm.com>
Subject: RE: Hierarchical vs. Parallel linkage
To:
"Richard Witkopp" <RWitkopp at phxa.com>
Cc: PowerHouse listserver
<powerh-l at lists.sowder.com>
Message-ID:
<OF23FDD4C1.837CAD0A-ON86257E00.00503F68-86257E00.0051AE2C at us.ibm.com>
Content-Type:
text/plain; charset="utf-8"
Hey Richard!
You are probably one of the last
people who remember my old MattO (Oh, not
zero) tag. :-)
I don't know if
there are still PH conventions, but I doubt it. Oddly
enough I gave a more
complete version of this many, many years ago at a
Cognos convention and it
was my biggest ego boost moment. The room not
only filled completely, but
there were people sitting on the floor at my
feet so they could see it!
I
wish I still had that original paper, as it seems to be a pretty
confusing
topic still. (Anyone have one in their archives?)
Also, If you have a QTP
with all those files, you might also be interested
in another paper I wrote
(years ago) called the "Care and Feeding of QTP".
I have a PDF of it still I
can send you if you'd like (I can't post it to
the list, if I recall).
However, that has more to do with output timing
than hierarchical vs. parallel
linkage.
My main advice for debugging Access statements has always been, take
the
Access and put it in Quiz and write a VERY basic Report statement that
dumps something from every files. Then go look at the files individually
and make sure they match what you see. If that works, they your problem
is
elsewhere. And honestly, with QTP, it is usually the "elsewhere" that
is the
problem. :-)
This has been fun!
Cheers,
Matt
From: "Richard
Witkopp" <RWitkopp at phxa.com>
To: "PowerHouse listserver"
<powerh-l at lists.sowder.com>
Date: 03/05/2015 06:29 PM
Subject: RE:
Hierarchical vs. Parallel linkage
Sent by:
powerh-l-bounces+mohmes=us.ibm.com at lists.sowder.com
Nice, not-a-zero!
Are there still Powerhoose conventions? This would make a good
presentation.
I need to re-read this thing about 4 times. I?ve got a qtp
with about 5
parallel legs and 3 hierarchical files down each leg and it
doesn?t do
what it looks like it should do.
From:
powerh-l-bounces+rwitkopp=phxa.com at lists.sowder.com
[
mailto:powerh-l-bounces+rwitkopp=phxa.com at lists.sowder.com] On Behalf Of
Matt Ohmes
Sent: Thursday, March 05, 2015 3:57 PM
To: PowerHouse
listserver
Subject: Hierarchical vs. Parallel linkage
Hi all,
I promised
I would write up something on the differences between
Hierarchical and
Parallel linkage in Quiz (and QTP), so here we go. (And
for those of you who
know all this already, please forgive my trip down
memory lane.)
Oh, and
I'm assuming no bugs in the product regarding syntax. :-)
Differences
between Hierarchical and Parallel linkage in Quiz (and QTP)
In all examples
below, we will use 3 files. Assume all are indexed files,
keyed on Emp_No
File Employees: 5 records;
Emp_No Emp_Name
1 Bob
2 Mary
3
Pat
4 Kim
5 Sean
File Paychecks: 5 records
Emp_No
Pay_Date Amt
1 2015-01-01 100 (Bob)
1 2015-01-15 200
(Bob)
2 2015-01-01 300 (Mary)
2 2015-01-15 400 (Mary)
3 2015-01-15 600 (Pat)
Note: 4 Kim and 5 Sean just started and
have not been paid yet)
File Dependents: 4 records;
Emp_No Dep_Name
1
Bob Jr (Bob's 1st dependent)
1 Bobbie (Bob's
2nd dependent)
2 Marie (Mary's dependent)
5 Shawn
(Sean's dependent)
Note: 3 Pat and 4 Kim do not have any dependents
This
will be my Report statement for ALL examples:
Report &
Emp_No of
Employees &
Emp_Name of Employees &
Pay_Date of Paychecks &
Amt of Paychecks &
Dep_Name of Dependents
First Hierarchical linkage:
Access Employees &
Link Emp_No of Employees to Emp_No of Paychecks &
Link Emp_No of Employees to Emp_No of Dependents
Emp_No Emp_Name
Pay_Date Amt Dep_Name
1 Bob 2015-01-01 100 Bob Jr
1 Bob 2015-01-01 100 Bobbie
1 Bob
2015-01-15 200 Bob Jr
1 Bob 2015-01-15 200 Bobbie
2
Mary 2015-01-01 300 Marie
2 Mary 2015-01-15 400
Marie
Obvious problems. Pat, Kim, and Sean are not showing up at all because
no
files are optional. I'm assuming the use of Optional in hierarchical
linkage is well understood so I won't go into that discussion here. This
post is about parallel linkage. :-)
A bigger problem with the report is
that paycheck and dependent records
are being duplicated. Hierarchical linkage
is simply wrong for this data;
the results make no sense.
Using
hierarchical linkage for this data is essentially saying, for each
paycheck an
employee has received, read down the list of all their
dependents. That makes
no sense. Just as it would make no sense to say,
for each dependent an
employee has, show me every time they have been
paid. The number of paychecks
I receive is based on how long I have
worked for my employeer, not how many
dependents I have. Similarly, the
number of dependents I have is not based on
how long I have worked for my
employer.
Paychecks and Dependents are
both, obviously related to Employees, but
they are NOT RELATED TO ONE ANOTHER.
Instead of one of these files "driving" the other, they should be read in
parallel; i.e. at the same time. THIS is why we have parallel linkage.
Parallel linkage is saying, for each employee, show me each of their
paychecks AND each of their dependents.
Let's try parallel linkage in the
Access statement and what that does to
our report results.
Access
Employees &
Link Emp_No of Employees to Emp_No of Paychecks &
AND
Emp_No of Employees to Emp_No of Dependents
Emp_No Emp_Name Pay_Date
Amt Dep_Name
1 Bob 2015-01-01 100 Bob Jr
1 Bob
2015-01-15 200 Bobbie
2 Mary 2015-01-01 300 Marie
2
Mary 2015-01-15 400
3 Pat 2015-01-15 600
5 Sean
Shawn
This makes more sense for this data. Each paycheck is reported once
and
each dependent is reported once.
Notice, as long as there is at least
one record FROM ANY of the parallel
files you get a record complex returned.
Parallel files are, more or less,
"semi-optional".
However, also notice
employee 4 Kim was not reported. This was because
Kim did not have any
paychecks (yet) or dependents. If you want to see Kim
on the report as well,
make ANY of the parallel files optional; it doesn't
make any difference which
one because of the "semi-optional" nature of
parallel files.
Access
Employees &
Link Emp_No of Employees to Emp_No of Paychecks &
AND
Emp_No of Employees to Emp_No of Dependents Optional
Emp_No Emp_Name
Pay_Date Amt Dep_Name
1 Bob 2015-01-01 100 Bob Jr
1 Bob 2015-01-15 200 Bobbie
2 Mary
2015-01-01 300 Marie
2 Mary 2015-01-15 400
3 Pat
2015-01-15 600
4 Kim
5 Sean Shawn
To be honest, parallel linkage is not that common. It is fair to ask, "If
two files aren't related to one another, why are you writing a report with
both of them on it?" But it does have its uses from time to time.
I
hope that explanation helps. (It was fun writing about good old Quiz
again
after all these years.)
Cheers,
Matt
securemail.phxa.com made the following
annotations
---------------------------------------------------------------------
NOTICE:
The information contained in this e-mail and
any attachments is confidential
and may be privileged
or otherwise protected from disclosure.This e-mail is
intended solely for the use of the named addressee.
Any other use, printing,
copying, disclosure or
dissemination may be subject to legal restriction. If
you are not the intended recipient, please contact the
sender and delete all
copies including any
attachments.
-----------------------------------------------------------------------
=
= = = = = = = = = = = = = = = = = = = = = = = = = = =
Mailing list:
powerh-l at lists.sowder.com
Subscribe: 'subscribe' in message body to
powerh-l-request at lists.sowder.com
Unsubscribe: 'unsubscribe <password>'
in message body to
powerh-l-request at lists.sowder.com
http://lists.sowder.com/mailman/listinfo/powerh-l
This
list is closed, thus to post to the list you must be a subscriber.
Add
'site:lists.sowder.com powerh-l' to your search terms to search the
list
archive at Google.
-------------- next part --------------
An HTML
attachment was scrubbed...
URL:
<http://lists.sowder.com/pipermail/powerh-l/attachments/20150306/8c610fe4/attachment.htm>
------------------------------
--
=
= = = = = = = = = = = = = = = = = = = = = = = = = = =
Mailing list:
powerh-l at lists.sowder.com
Subscribe: "subscribe" in message body to
powerh-l-request at lists.sowder.com
Unsubscribe: "unsubscribe
<password>" in message body to
powerh-l-request at lists.sowder.com
http://lists.sowder.com/mailman/listinfo/powerh-l
This
list is closed, thus to post to the list you must be a subscriber.
Add
'site:lists.sowder.com powerh-l' to your search terms to search the list archive
at Google.
End of powerh-l Digest, Vol 113, Issue
15
*****************************************
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sowder.com/pipermail/powerh-l/attachments/20150306/a531139a/attachment-0001.htm>
More information about the powerh-l
mailing list