Idea: Filtering - Show only Named Query subset values?

April 2, 2013 7:09AM

Hi all - When using a Named Query to define the visible Subset of the data in one or more views in Omniscope DataExplorer, the Sidebar filter devices are not tracking only the viewable data on the tab. The filtering options are still offering values that are filtered out by the Subset settings of the views on that tab?

Also the auto-complete filtering suggests values that are not valid for that tab as these values are not part of the selected subsets and queries applied on the views of the tab.

I would like to set up the dashboards such that the Sidebar filter devices are based only on the intersection of all named queries applied to views on the tab. Is there a setting that I missed or is this not possible (yet)?

steve · April 3, 2013 5:35AM

Alex- No, you haven't missed a setting. The filter devices don't reflect the data Subset choices in your views.

But it's easy to set up. You should be able to show the Queries filter device on the Sidebar in each tab and tick all the relevant queries (for Intersection).

If you want Union, you can tick each query, one at a time, and add each to a new, aggregate union query.

Mees · April 4, 2013 11:19AM

Hi Steve, thanks for this. We have tried this, but it doesn't solve our issue yet.
We have iok files with several tabs, each tab is based on it's own subtable within the Omniscope table. There are columns that are used for several subtables; I have attached an example.

In my example, tab2 makes use of the green subtable. When column C is a text field, it shows values in it's droplist that belong to the yellow subtable also. However, these are values that belong to tab1 and shouldn't be part of the droplist in tab2.

How can we hide these values from the droplist in the text field

Thanks
A

tjbate · April 10, 2013 10:46AM

Arjan - If you are appending un-related data sets, but some of the field names are exactly the same, Omniscope will align all the fields with exactly the same names and assume that all values in the field are valid from a filtering perspective. If you do not want this, branch the flow upstream into divided subsets, and change the field names in each subset in DataManager slightly, so that the fields will not be aligned when you append all the different data sets together again prior to loading all the appended rows into the Omniscope in-memory block.

Mees · April 10, 2013 10:58AM

Thanks Tom,

We are aware of this, but we re-use column names on purpose in order to reduce the number of columns. As Omniscope is based on a one table model, this is a way of having multiple tables in one big table, and keep the number of columns relatively low. As we have massive files, adding more columns as you describe is not an option as it has negative impact on performance. In that case we should use a new iok file and that is not what we want.

Is there any way that the droplist values are related to named queries?

tjbate · April 10, 2013 11:05AM

Arjan - Not at the moment, but we understand the need... I have re-classified this as an Idea, and we will think about it a bit more...

Mees · April 10, 2013 11:07AM

Great thanks, in case you need input give me a call. we use this a lot so always willing to help

steve · April 11, 2013 5:00AM

If you simply want to customise which fields are available in each drop down, click the tiny spanner/wrench icon at the top of the menu. This is currently configurable on a per-tab + per-dropdown-type basis.

tjbate · April 11, 2013 6:39AM

Arjan - If the problem relates to illogical values shown in the Sidebar as filter categories for an appended field, make sure that the option 'Show values excluded by other devices' is NOT ticked on the relevant filter device drop-down. This makes the Sidebar filter device aware that the illogical/inapplicable values have already been excluded by the Named Queries.

Welcome!

Categories

Ideas Parade

Tagged