09-01-2017 03:54 AM
I set up FEATURE DATA and BOUNDARY DATA (see pic 01).
Then, the relation between tables is fixed: TAB_DETAILS.ID_PARENT -> TAB_BOUNDARY.ID and aggregation function is selected: Boundary statistic - Feature Count. The pie chart is rendered properly and the hint shows number of features (see pic 02).
When I switch Boundary statistic function to Sum or to Average with indication of summary Field name, the chart and the hint don't follow this change (see pic 03). I can't get other dimension than feature count.
Solved! Go to Solution.
09-26-2017 08:08 AM
This first chart will continue to follow the specified theme field with tooltip equal to feature count as the boundary statistic is pertaining to the boundary data and the chart to the specified theme. To confirm this you may change color of boundary and observe results in the boundary features when choosing AVG and then SUM.
To see an update on a chart a second chart should be added using AVG/SUM measure as I can see has been done in stat03.png. it appears that you are handling this correctly only that the first chart is not tied to the 'boundary statistic'.
09-27-2017 04:42 AM
Let's start from the begining. For the purpose of better communication I labeled configuration tabs with letters (A, B, C, ...)
1. Tab A. Selection of Boundary Data and Feature Data.
2. Tab B. Relationship between Feature Data and Boundary Data, selection of Main Theme, and Boundary Statitics:
3. Tab D. Definition of Additional Theme:
4. Tab E. Definition of Additional Chart:
My doubts concerns configuration that is beeing set on tab B, and its impact on Chart presenting Main theme and all Additional themes.
a) Shouldn't I be able to see results of Sum or Avg functions on this chart? If not, what do fields B9 and B10 serve for?
b) There is only one place (field B8) where I can select field from Boundary Data. I guess it is used to define the relation between Boundary Data and Feature Data. Is that by design?
09-27-2017 07:04 AM
To answer your questions:
A) No, fields b9 and b10 purpose are attempted to be explained below B) Yes
Likewise I will use some screenshots to explain this answer and use your tab/field marking hopefully to help. Please note that this data is of course different from yours, but illustrates a use for Analyzer View in this manner.
Here I have a feature input and boundary data
If I set up only a ‘theme field’ (tab b) then the chart will display based on the feature count of that field.
The boundary data does not change and the feature data changes now only to be the color (tab c) specified on the color tab as it has now become activated.
If I now attempt to color the Boundary Data with a number of slices the whole feature set is colored the same since there is no statistic specified yet to determine the slices (classification) which would determine the boundary feature color.
To manipulate/analyze the boundary feature in this theme I return to the Theme tab (Tab b) and define the relationship between input feature and boundary feature (b8) and I leave the boundary statistic (b9) as the default ‘Feature Count’
When I choose ‘Apply’ the result is that the boundary data is now colored based on the boundary statistic aggregation. There is no change to the chart showing the theme as the theme as not changed.
Subsequently I can change the boundary statistic (b9) to be avg or sum and note the change in boundary feature coloration based on the result. Still no change to the chart for theme should be expected as it is not affected.
Sum Sentiment (Not very useful in regards to this dataset, but nonetheless a change is seen vs Average)
Should I wish to see a chart that represents the Average Sentiment I would need to use chart tab (tab f) specifying the field and statistic, etc. Following along to the previous, I would not expect this chart setting to have an effect on the boundary statistic and thus boundary coloration as that has not been changed by this process.
To bring both in line I would need to set boundary statistic back to Average Sentiment on theme tab (tab b)
I hope that makes sense and helps explain the tabs and fields specific to your inquiry.