hatman
Well-known Member
- Joined
- Apr 8, 2005
- Messages
- 2,664
I couldn't manage without pivot tables!!!
Maybe I'm doing something wrong but I tried summarising some pretty extensive data (50,000 employees worth of pay data for a year) on Monday using SUMPRODUCT and the formula took all afternoon to calculate for just one month (actually I went home and left it going so not sure how long it took).
I then had a moment of clarity during the night and realised how to do it using pivots and I had the other eleven months summarised an hour and a half after getting to work the following morning.
Learn them, they're legendary :wink: !!!
Yes, I see how a pivot-table is perfect for that application...
The only challenge from my personal point of view is that I would need to fabricate data that lends itself to a pivot-table structure... and that almost defeats the purpose of the lesson. I would like to add them to my arsenal, but I have yet to come across data in my daily duties where that type of data drill-down would even be remotely useful.
Just to give a sample of my most data intensive workbook:
This is a sample from a table with 328 members, representing the coded part numbers for electrical connectors on a piece of Space Station hardware:
Node_3_Electrical_Tool_Crossref_ver_1-1.xls | |||||||||
---|---|---|---|---|---|---|---|---|---|
A | B | C | D | E | F | G | |||
1 | ORUName | ORUPartNumber | ItemSchematicID | PartNumberofItemORDrawingWherePinoutSchematicisLocated | ConnectorID | ConnectorPartNumber | |||
2 | WPA | AAA | A107 | J1 | NATC00T11N98PB | ||||
3 | WPA | AAA | A107 | J2 | NATC00T11N35PN | ||||
4 | WPA | ASD | A108 | J1 | NATC07T15N35PN | ||||
5 | WPA | CatalyticReactor | SV825455-1 | 0229 | SV826686-1 | J1 | NATC00T11N35PA | ||
6 | WPA | CatalyticReactor | SV825455-1 | 0231 | SV825503-2 | J1 | NATC00T11N35PN | ||
7 | WPA | CatalyticReactor | SV825455-1 | 1126-1 | SV825457-1 | J1 | NATC00T11N35PN | ||
8 | WPA | CatalyticReactor | SV825455-1 | 7047-1 | SV825455 | J3 | NATC00T13N35PN | ||
Hardware |
And here is a sample from one of 5 other tables on 5 sepearate sheets that have 50-250 members... this particular table represents parts numbers for Test Cables, with the part numbers for the electrical connectors on each end:
Node_3_Electrical_Tool_Crossref_ver_1-1.xls | |||||||||
---|---|---|---|---|---|---|---|---|---|
A | B | C | D | E | F | G | |||
1 | ConnectorID(itemside) | ConnectorPartNumber(itemside) | TestToolPartNumber | TestToolDescription | ConnectorID(out) | ConnectorPartNumber(out) | Notes(fortoolorconnector) | ||
2 | P1 | D38999/26FD35SN | SV825464CT206 | ReactorHealthHarness#1 | P1TB | D38999/26FD35PN | Valves | ||
3 | VLV4 | D38999/26FB35SA | SV825464CT207 | CatReactorO2ManifoldCable | PRIG | PT06A-10-6S | |||
4 | VLV5 | D38999/26FB35SN | SV825464CT207 | CatReactorO2ManifoldCable | PRIG | PT06A-10-6S | |||
5 | P1 | D38999/26FD35PN | SV825464CT208 | CatReactorO2ManifoldCable#2 | PRIG | PT06A-10-6S | |||
6 | P1 | D38999/26FB98SA | SV825487CT208 | GLSHarness#1 | P1TB | D38999/26FB98PA | Heaters | ||
7 | P2 | D38999/26FB98SN | SV825487CT209 | GLSHarness#2 | P2TB | D38999/26FB35PN | |||
8 | P3 | D38999/26FC35SN | SV825487CT210 | GLSHarness#3 | P3TB | D38999/26FC35PN | (9)rtdSignalsandLiquidSensor | ||
9 | P1 | D38999/26FE35SN | SV825502CT202 | WaterStorageHarness | P1-TB | D38999/26FE35PN | Don'tUseonWasteWaterORU | ||
Cable |
Now, how would a pivot table tell me that the test tool listed in row 3 of table 2 is one of several tools in that table that will connect with the hardware listed in row 5 of table 1? Once I know that I need to find all tools that have a D38999/26FB35SA connector listed in column B or F, an autofilter is almost the right tool... except that I can't get the union of matches in both columns. I ended up doing this with a whole slew of Match() and Index() formulas, that look like this: =MATCH(B3,INDIRECT($C$3&"!B1"):INDIRECT($C$3&"!B"&$A$1),0) and =MATCH($B$3,INDIRECT($C$3&"!F1"):INDIRECT($C$3&"!F"&$A$1),0) so I can obtain matches in BOTH columns, where C3 contains the name of the sheet (and yes, a year later, I DO, in fact see how I could make these formulas a little bit more compact, but I just don't feel like taking the time to make the changes in all 50 rows by 15 columns in each of 7 sheets, even though I might see a noticeable gain in calculation speed). I defy you to demonstrate how a Pivot Table would be more applicable in this situation.
I grant you that pivot tables are amazingly powerful for drill-down of large data sets... but I guess I'm still stuck on the fact that if you don't need to do data drill downs, and/or your data sets are TINY, a pivot-table is at best too big of a hammer, at worst it just isn't applicable.
I do, however, appreciate your opinion, and as I said to greg: now that I have a better appreciation for what they are good for, I will be keeping pivot tables in reserve for a time when I may find them useful.