You just need to take Gianni's answer and apply the details that you didn't include in your original post. Filter Column B on RANK(COUNT( "Col B" by "Col A"))=1 or TOPN(COUNT( "Col B" by "Col A"),1)

That was answered here: Doc ID 2067123.1 for 11G. I don't know a way to do it in the front end. Do you have the correct locale selected? That will change your default date formats.

You're talking about row-level security. Search on that and you'll find a lot of answers that are better than I can provide. Here's one from Gianni: https://community.oracle.com/message/12991195#12991195

Hello SonPat, I think your best hope of doing that is to pre-calculate the fields in the rpd. Good luck! Jerry

A simple two column example, top query has several missing dates, the bottom contains all Saturday dates: Weekly Delta formula: (sum(saw_1 by saw_0)-(MSUM(sum(saw_1 by saw_0),2)-sum(saw_1 by saw_0)))/sum(rcount(saw_1 by saw_0) by saw_0) This is a pivot table with Amount duplicated, shown as running sum:

By wrapping each Amount in an ifnull statement, this worked:

I'm still unclear about the issue. You can make a filter like: "Table"."Created Date" <=(CURRENT_DATE) AND "Table"."Created Date" > "START DATE" AND DAYOFWEEK("Table"."Created Date")=7 To get the difference between adjacent rows, a formula like "Col X"-(MSUM("Col X",2)-"Col X") works. (moving sum of last 2 rows minus the…

Hi, You can filter on hierarchical columns using selection steps.

Do you have the Print Rows Option for the view set to "All"?

Hi SonPat, Have you tried a union query where you add a query that that fills the skipped dates? It would have zeroes in the measure columns, and you would build the graphed columns at the header level using "saw_x" terminology. (Add a result column)

for what it's worth: The only time I've seen a problem like this was when a custom cs style of "display:none;" had been applied to a column header, but not to the column values. It doesn't sound related because you say the problem is intermittent. Any chance a conditional format of this kind was mis-applied?

I'm not familiar with the use of DDR,, but this solution works on dashboard queries: The value of a multi-select string prompt looks like: 00001,00002,00003 So, you have to do some parsing and replacement. In an analog to your example, I have a similar prompt for County. I use a regular expression to search the "Project…

Glad it helped! The only drawback to this method is that when you have to edit the query you need to temporarily replace the "XXX" with a valid value.

Hi Dennis, If you are using a presentation variable instead of "is prompted" in the query filter, you can use a dummy default value that produces no results. it may look like "Table"."Required Field"='@{varname}{XXX}' In the "No Results" window you can insert a message like "Complete the required prompts"

The simplest way is to duplicate the column in your pivot table, select the edit option on the duplicate,, and select "Display value as percent of column". You can edit the header of that column to read "Month Pct", or whatever.

It looks like you do have at least one null record in C1, but I'm not sure why that would prevent the conditional format from working. I just went back and confirmed what I described above, so I know it works in my example. Can you try it with a filter on C1 IN ('test', 'test1')?

Hi Chris, I adapted this from an earlier answer of mine: Custom groups are able to accept conditional formatting. In the pivot table below, there are two groups. The first groups the 3 values of "Status" into a group named "Totals". I have placed a conditional format on the measures that when "Status" is null, use a bold…

Hi Dravel, This is a stacked line-bar pivot table and chart that performs as you've described. The key is that the calculated percentage column, in this case "Consultant Hrs Pct" is showing the correct amount prior to producing the chart. You may need to produce 3 calculated columns: a "Yes Volume" column, a "No Volume"…

Hi Steel, I don't have much experience working with future record dates, but it should just require variations on the logic I've used above. You want to exclude all results prior to the current date, without any filters. As far as I can see, that would require a flag column, such as CASE WHEN date < CURRENT_DATE THEN 'B'…

What does your query filter look like? You can filter the query with this, or something like it: "Contract"."Award Date" >TIMESTAMPADD(SQL_TSI_WEEK,-12,current_date)