I have 2 weeks of data and I am trying to present summary (by week commencing) of all metrics (CPC, Clicks etc) horizontally. I also need to present differences between those two weeks horizontally. Table view - good but I can't show differences as summary row doesnt have option (deduct) Pivot view - is good to show the differences but can't show all metrics in one view as they are not categories; it also doesn't let me chose "week commencing" because it is a date so the split is impossible.
As discussed, please aggregate by week commencing and de-pivot to get all your metrics in one column. You can then use pivot view. Create a copy of the date field and convert it into a category field.
Darek - Horizontal layout by weeks as columns is typical in spreadsheets, and you can deliver your data set oriented that way, but analysis and calculation of time series data in Omniscope is always best done vertically, using the De-Pivot data operation to change the orientation and Pivot operation to change it back if you have to deliver spreadsheet output in horizontal 'weekly' columns. Note that De-Pivot and Pivot operations are not related to what Excel calls a Pivot Table, which is a form of aggregation by specific fields/columns presented in a two-dimensional table...same as the Omniscope Pivot View.
Thank you. If I didn't have to calculate MOM/YOY changes then I would apply this what you are suggesting. But this is fine my question was about colouring data by a single row. Is it possible?
Dariusz - Not in the Table View....colouring in the Table View is based on columns/fields setting, with row-by-row variation settings to aid legibility/visibility vertically, not horizontally. More powerful conditional formatting options are available in HTML table layouts in the Content View.
The tab/page layout with horizontal timeslice orientation you are working with in this file may have been specified as a data delivery format to facilitate re-use, but horizontal rows of numbers are not easy to follow visually.
What you are looking for is something like 'sparklines', small horizontal lines that show the ups and downs in the horizontal time series row values, which are much more understandable than scanning horizontally across rows of figures, coloured or not.