You are not logged in.

Currency Conversion 2.0

Dear visitor, welcome to Palo Community Forum. If this is your first visit here, please read the Help. It explains in detail how this page works. To use all features of this page, you should consider registering. Please use the registration form, to register here or read more information about the registration process. If you are already registered, please login here.

  • 1
  • 2

h_decker

Sage

  • "h_decker" is male

Posts: 763

Date of registration: Oct 26th 2005

Location: Cologne, Germany

  • Send private message

21

Monday, October 22nd 2007, 3:04pm

Hi, I already reported it. Maybe you can try to export your data and reimport it with PALO.SETDATA and a trim function or.....maybe this is the most comfortable way: Add a trim-function to your rule like:

['TEST'] = N:['Unit Price (LOC)'] * palo.data("Revenue_Budget","Budget Currencies",!'FY',TRIM(palo.data("Revenue_Budget","#_Regions","Country Budget Currency",!'Regions')))


Holger

alex_s

Professional

  • "alex_s" is male
  • "alex_s" started this thread

Posts: 31

Date of registration: Mar 2nd 2006

  • Send private message

22

Tuesday, October 23rd 2007, 9:05am

Hi,

Just went ahead and made the changes manually (once I got the rhythm going it went fairly quickly).

However, even after deleting the trailing space in the region attribute fields, I still get NULL values when I try to do the "join". It returns the correct currency code for each country, and when I hard code "EUR" it returns the EURO FX rate, but, when I embed one in the other, it just gives NULL. Frustrating.

Guess I'll just start with a fresh DB and pay attention to the "hidden space" issue.

Thanks again for your help.

-Alex

h_decker

Sage

  • "h_decker" is male

Posts: 763

Date of registration: Oct 26th 2005

Location: Cologne, Germany

  • Send private message

23

Tuesday, October 23rd 2007, 9:56am

I noticed that in one dimension (I think it was Currency_Codes) you got attributes elements equal to the original element names. I deleted the attribute elements. Please try that and check again. I also can attach my modified db if you like.

Holger
h_decker has attached the following file:

This post has been edited 1 times, last edit by "h_decker" (Oct 23rd 2007, 10:00am)

alex_s

Professional

  • "alex_s" is male
  • "alex_s" started this thread

Posts: 31

Date of registration: Mar 2nd 2006

  • Send private message

24

Thursday, October 25th 2007, 11:20am

Hi Holger,

Wasn't able to get it working with your revised DB (the one you attached to the last reply), but I did mange to rebuild from scratch and now it seems to be working fine - hurray!

Thanks again for your help and persistence. Now that it's working, it's really great - so nice to be able to deal with multiple currencies (and more generally to do this sort of cross-cube calling).

best regards,

Alex

h_decker

Sage

  • "h_decker" is male

Posts: 763

Date of registration: Oct 26th 2005

Location: Cologne, Germany

  • Send private message

25

Thursday, October 25th 2007, 12:15pm

Hi Alex,

nice to hear.

Holger

Pommie

Sage

  • "Pommie" is male

Posts: 411

Date of registration: Feb 21st 2007

Location: Lytham UK

Occupation: OLAP consultant specialising in retail

  • Send private message

26

Friday, October 26th 2007, 1:30pm

It would be great if Palo could be as tolerant regarding spaces as "other similar products". The one I work with now ignores additional or missing spaces. It is often a helpful feature and I have never had it cause me a problem


Best wishes

Pommie
Best wishes

John Hobson
The Planning Factory, Lytham, UK
www.planfact.co.uk

h_decker

Sage

  • "h_decker" is male

Posts: 763

Date of registration: Oct 26th 2005

Location: Cologne, Germany

  • Send private message

27

Friday, October 26th 2007, 1:43pm

Maybe you can post it as feature request in bug tracking system?

Thanx in advance
Holger

This post has been edited 1 times, last edit by "h_decker" (Oct 26th 2007, 1:43pm)

alex_s

Professional

  • "alex_s" is male
  • "alex_s" started this thread

Posts: 31

Date of registration: Mar 2nd 2006

  • Send private message

28

Monday, October 29th 2007, 8:40pm

calculation speed/performance

hi holger,

FX conversions are now working fine. Thanks again for your help finding the problem with the extra spaces!

i've also managed to follow the examples in the manual to upload the data (will define different templates next time to make this easier).

the issue I'm facing now is that the cube seems to take quite some time to refresh. i assume this is due to the calculation of the rules (and perhaps also that i'm running on my local machine).

are there any guidelines about how to maximize efficiency for calculations (e.g., splitting the calculated metrics into different cubes, for example, to have one for reporting and one for data entry ... ?).

also, are there any guidelines in terms of how hardware upgrades (more ram, running on a server with a more powerful CPU) might help improve performance.

thanks in advance,

alex

h_decker

Sage

  • "h_decker" is male

Posts: 763

Date of registration: Oct 26th 2005

Location: Cologne, Germany

  • Send private message

29

Tuesday, October 30th 2007, 12:37pm

RE: calculation speed/performance

Hi,

>also, are there any guidelines in terms of how hardware upgrades (more ram, >running on a server with a more powerful CPU) might help improve >performance.

I think there are some posts concerning that. More ram is always a good thing because PALO holds data in memory. Just browse forum for more info.

Concerning the performance with your rules, I recognized that the same for my testmodel: Its the EUR Rule that slows down the slices with EUR-Values. I already added performance enhancement concerning such kind of rules to the PALO wishlist. So let us see what happens.

One more thing about our "EUR " instead of "EUR" issue. That is not a PALO bug. We can give a big THANK YOU! to Excel for that. Once you entered "EUR " in a column and enter e.g. "EUR" in the next row, Excel toggles its auto completion and you got "EUR " instead. So we just must have a closer look at what Excel does with our entries......as always :-).

Holger

alex_s

Professional

  • "alex_s" is male
  • "alex_s" started this thread

Posts: 31

Date of registration: Mar 2nd 2006

  • Send private message

30

Thursday, November 1st 2007, 4:47pm

RE: calculation speed/performance

Hi Holger,

Yes assumed more ram would help - so am looking into this ... am also testing db on one of our servers instead of my localhost. will keep you posted.

but yes, optimization of rules engine for enhanced performance would be nice (might also be worthwhile to offer some tips on db modeling and rule writing best practices to make rules more efficient given the existing engine.)

and also glad (and not terribly surprised) to hear the bug was in excel not palo :-)

will pay careful attention to how Excel "adjusts" my entries ...

thanks,

alex

Pommie

Sage

  • "Pommie" is male

Posts: 411

Date of registration: Feb 21st 2007

Location: Lytham UK

Occupation: OLAP consultant specialising in retail

  • Send private message

31

Tuesday, November 6th 2007, 7:11pm

Is there an intention to create "accelerators" or "feeders" to speed calculations?
Best wishes

John Hobson
The Planning Factory, Lytham, UK
www.planfact.co.uk

Stephanie

Sage

Posts: 540

Date of registration: Oct 23rd 2006

Location: Freiburg

  • Send private message

32

Wednesday, November 7th 2007, 7:39pm

Hello,

We are continuously working on to speed up Palo performance.
Depending on your application it could help to adapt the caching
parameters in your palo.ini file (see your local palo.ini.sample).

Features like local data caching and server subsets are on
the Palo Roadmap for Palo 2.1 upcoming November.

Regards,
Stephanie

alex_s

Professional

  • "alex_s" is male
  • "alex_s" started this thread

Posts: 31

Date of registration: Mar 2nd 2006

  • Send private message

33

Friday, November 9th 2007, 3:25pm

Hi Stephanie,

Any suggestions on how best to tweak ini file (or other application/machine parameters) to improve performance.

My ini file has a cache barrier set to 5000 (guess this is default). what would you recommend.

I've ordered a PC with more ram (4gb) and a more powerful cpu (2.4ghz dual core) so i would hope that will solve the performance issues. This is key, since we are trying to use it for "Real-time" budgeting discussions, but the response time is way too slow (several minutes for a refresh ...)

Does DB size or fill ratio impact performance? (I have 8 dimensions, 444517632 cells and 9175 filled cells (0% fill ratio). This is linked to the currency cube, which has 2 dimensions with 18 cells at 100% fill ratio.)

Any other settings that I should explore to try to boost response time? Any design suggestions that might do the same?

Thanks!

-Alex

vesi

Sage

  • "vesi" is male

Posts: 97

Date of registration: May 11th 2006

Location: Czech Republic

  • Send private message

34

Tuesday, June 24th 2008, 12:14pm

Multi Currency Conversion

Hi all,
I have cube with dimensions:

CENTRE
CURRENCY
CALCULATE
VERSION
YEAR
PERIOD
INDCF

In dimension CURRENCY I have:

CU1 [CZK]
CU2 [EUR]
CU3 [USD]
CU4 [SK]
CU5 [PZ]

and in dimenzion CALCULATE:

CUR [value in currency]
VAL [calculate value]

my rules:

['CU1','VAL'] = N: (['CCtotalX','CU2','CUR','CURR'] * ['CU2','CUR']) + (['CCtotalX''CU3','CUR','CURR'] * ['CU3','CUR']) + (['CCtotalX','CU4','CUR','CURR'] * ['CU4','CUR']) + (['CCtotalX','CU5','CUR','CURR'] * ['CU5','CUR']) + ['CU1','CUR']

['CU2','VAL'] = N:['CU1','VAL'] / ['CCtotalX','CU2','CUR','CURR']

atc.

where:
['CCtotalX','CU2','CUR','CURR'] = rate of exchange
['CU1','CUR'] = value in currency


If I change element CUR/VAL I have recalculate values from all currency to selected currency :P

Sorry for my english :rolleyes:

This post has been edited 6 times, last edit by "vesi" (Jun 24th 2008, 3:31pm)

  • 1
  • 2
Rate this thread