Recent content by nagrajm

Michelle Shop around for third party schedulars which have lot more features and can be installed on a server. We have, for example used apps like TechScheduler and Tivoli Maestro to schedule daily loads and report runs. HTH

Here's a piece of sql that will help you find the date of the first monday of a given year and month. TO_DATE(CONCAT(CONCAT(?Year?,?Month?),TO_CHAR(TRUNC(DECODE(TO_NUMBER(TO_CHAR(TO_DATE(CONCAT(CONCAT(?Year?,?Month?),'01') ,'YYYYMMDD' ) ,'D')),1,2 ,2,1,3,7,4,6,5,5,6,4,3) ) )) ,'YYYYMMDD' ) In...

Is the date string in your database in the format YYYYMMDD because what the others are trying to tell you is that unless the date sting is of YYYYMMDD format you cannot determine what day of week the record pertains to. If the date string in your database is missing the DD you cannot find the...

Add the following filter condition; Start_Date_Time between to_date(concat(date-to-string(?Prompt?),' 08:00'),'DD/MM/YYYY HH24:MI') to to_date(concat(date-to-string(add-days(?Prompt?,1)),' 07:59'),'DD/MM/YYYY HH24:MI') I have tested the code and it works, but for any reason if this gives an...

In the property sheet of the Count measure, select the Time State Rollup of the Last Period or the Current period to show the count at "a point in time" rather than the numbers rolled up (Summed).

I do not know about the Curdate() function but year(Now()) and month(Now()) should strip away the year and month. This always works.

You can by creating the column definition again. Associate each of them to two different groups to get two different summary numbers.

Here's one interesting solution. Try Count = Char-Length(Replace(Days_on_Contract,"N","")) In fact this is very similar to what you did in Excel except that you had to suppress N instead of Y to shorten the formula.

Mark...You are missing the closing bracket at the end of the monthname function calculation.

In a case like yours, I generally use the left and right sides of the report to show the summary and the detail. I do not use the top and bottom approach precisely for the reason that you are not able to. It's a lot of work formatting the detail and summary on the left and right but it should...

Pr07 Your method is one way to do it which I don't think should be inefficient, if that is the only way to accomplish the result in impromptu? How long does it take? Another method would be to create a database view (with the same logic that you would in creating the impromptu report) of all...

Do not allow other users, except you, to log in as creator.

No. If you are using impromptu with Oracle, for example, use the Week() function on the date to extract the week no. WEEK_NO = WEEK(SALE_DATE) Use the WEEK_NO column in the transform as "Label Colum" in the property sheet for Week Level in the date dimension. I hope it is clear.

In order to create a label for weeks from 1-52, you will have to create a calculation in the query to convert a date to it's corresponding week no. Use this calculated column for "Label Column" in the properties for Level - Week in the date dimension in transformer. HTH.

Another method, without having to use a dataset, would be to pull a report of all transactions for the period (TARGET DATE minus ONE YEAR) and then use the summary filter to weed out customers who had transactions before the target date. In order to accomplish that, group the report on...