Recent content by MattOh

The "value if null" option on the element properties is only applied on output so you could use it to replace null business keys, but only in a pre-pass build. It happens too late to help in the build where you are checking your lookup. In DS 7.0 you should be able to create a...

Hi Xantix, If you really want to do this, you can turn on template access by going to the Dimensions in the Library, select the Lookup you are useing in your fact build, open the property sheet, and select the Data Access tab. There is a checkbox that lets you use template access. However, I...

Hi Gurudatta, What you are trying is not that uncommon. It's often used in demographic analysis and is also called "banding". Other examples include age ranges and income ranges. See pp 155-157 in Kimballs, Data Warehouse Toolkit (Second Edition) for more details on demographic (or...

Hi Blom, I'll try to avoid re-thrashing the aggregate table discussion again too. But it might creep back in. :-) My comment about the BI value of Booz's original table was based on the fact that all you could apparently get out of it was the number of meetings held each month. Nothing about...

Hi Booz, While hundreds of thousands of meetings a year may sound like a lot, in DW terms, it's still pretty small. :-) Don't assume the numbers will be a problem as long as you aren't severly limited in processing power or storage space. As far as concept of Dimensions go, don't think of...

Hi Booz, What you have just created is a VERY simple, periodic snapshot table. Technically, there is nothing else you need to put in it if all you want to know is the total number of meetings in a month, for all sites, for all attendees. But I have to wonder how much useful Business...

Hi Loukas, Me again. :-) I'm presuming you want to FTP the file from the remote machine to your local server before processing it in SQL TXT. If that is the case, I would suggest you set up a JobStream. Make the first node in the JobStream a Procedure node. In that node, you can call any...

OK, now on the Materials tables. You are reading two data sources in, merging them(?) and delivering to one target dimension table? Only if the dates don't match, you want a new row written out to the table with a new Materials_ID? This sounds similar to the first problem but with a twist...

Let's take the Customer one first. I'll handle the second question in a separate post. In Customer_LF and Customer_SM, is Customer_Code the key in your source system? I'm guessing Customer_ID doesn't even exist in your source system tables, is that right? If so, you don't even need slowly...

Hi Loukas, I can answer your questions, but I want a little more clarity on what you want to do. Both of these questions sound a awful lot like Type 2 Slowly Changing Dimension support, which DecisionStream can handle with build in options. But just to confirm... First, if you are creating a...

Hi TIA, Actually, I do have a suggestion for dates and nulls, but it's going to sound funny. Use characters. As you have pointed out, putting dummy dates into a date field is problematical. You typically can't put in "0" (which is illegal) and dates like 31-dec-4712 have their own...

I'm not sure what OWB is or what features it supports, but if possible, I would recommend replacing nulls with some known, but not normally available value before you load them to your dimension table (e.g. "?" or "N/A"). You have to be very careful when testing for nulls...

Hi Loukes, That message usually means that you don't have a proper authorization key for DecisionStream. Without the key, you can design builds but not execute them. Contact Cognos for a license key. I believe you can get a 30 day evaluation copy of DecisionStream at no cost on their web...

Perhaps you should look at a different tool. Check out PowerPlay and Impromptu by Cognos. Everything you describe is very common and the users seem pretty comfortable with it. If you can't change BI tools then SQL access directly against aggregate tables may indeed be your best option. Sorry. Matt

Hi T, I agree, data warehousing is not easy or inexpensive. And if you want to get your users thinking of alternatives to direct SQL reporting from OLTP, then a flashy demo of a good OLAP tool, with the users' data, is a good way to do it. As for your question on standard reporting (e.g...