Data Expression to Reshape Attribute Table Removes Ability to Filter Map

Joshua_Flickinger

Background

I have a feature service in this format:

Trail_Name	Group1_Inspection	Group2_Inspection	Group3_Inspection
Trail 1	Not Inspected	Inspected	Inspected
Trail 2	Inspected	Inspected	Inspected

 

I needed to re-arrange the data for graphing in a dashboard, so I figured out how to use a Data Expression to reshape the attribute table into this format:

Trail_Name	Inspection_Group	Inspection_Status
Trail 1	Group1	Not Inspected
Trail 2	Group1	Inspected
Trail 1	Group2	Inspected

And so on...

Essentially, this way I could graph by Inspection Group and split the data in the graph by Inspection Status so I could display in the dashboard how many groups have completed their various inspections.  Assuming the method to the madness matters, you can see essentially how I did this in the code snippet at the end of the post.  The code works, and I'm able to create a graph the way I wanted.  However...

Question

The issue I'm having is that my graph does not allow me the option to filter my map, based on column selection.  When I try to configure the chart properties for filtering, it wants me to specify destination and target fields.  Well, I retained trail name in an attempt to make this work, but it doesn't allow that as an option.  So, I'm trying to figure out whether that was due to the way I split up my data through my data expression - or alternatively if it's just not possible after the data re-arrange.  Anybody have any clues?

Thanks.

Code Appendix

var portal = Portal('https://www.arcgis.com/');
var fs = FeatureSetByPortalItem(
    portal,
    '1234567890',
    0,
    [
        'TRLNAME',
        'Group1_Inspection',
        'Group2_Inspection',
        'Group3_Inspection'
    ],
    false
);

var g1 = GroupBy(fs, ['Group1_Inspection', 'TRLNAME'], [{name: 'Count', expression: "1", statistic: 'COUNT'}])
var g2 = GroupBy(fs, ['Group2_Inspection', 'TRLNAME'], [{name: 'Count', expression: "1", statistic: 'COUNT'}])
var g3 = GroupBy(fs, ['Group3_Inspection', 'TRLNAME'], [{name: 'Count', expression: "1", statistic: 'COUNT'}])

var features = [];
var feat;

for (var g in g1) {
    feat = {
        attributes: {
            TRLNAME: g['TRLNAME'],
            inspection_group: 'Group1',
            inspection_status: g['Group1_Inspection'],
            Combined_Segment_count: g['Count']
        }
    };
    Push(features,feat);
}

for (var g in g2) {
    feat = {
        attributes: {
            TRLNAME: g['TRLNAME'],
            inspection_group: 'Group2',
            inspection_status: g['Group2_Inspection'],
            Combined_Segment_count: g['Count']
        }
    };
    Push(features,feat);
}

for (var g in g3) {
    feat = {
        attributes: {
            TRLNAME: g['TRLNAME'],
            inspection_group: 'Group3',
            inspection_status: g['Group3_Inspection'],
            Combined_Segment_count: g['Count']
        }
    };
    Push(features,feat);
}

var fsDict = {
    'fields': [
        {'name':'TRLNAME', 'type':'esriFieldTypeString'},
        {'name':'Inspection_Group', 'type':'esriFieldTypeString'},
        {'name':'Inspection_Status', 'type':'esriFieldTypeString'},
        {'name':'Combined_Segment_Count', 'type':'esriFieldTypeInteger'}
    ],
    'geometryType': '',
    'features': features,
}

return FeatureSet(Text(fsDict));

jcarlson

Glad you figured it out!

The key to that second issues is this:

geometryType: ''

Without including the geometry of your features, there's nothing for the map to act against.

You can include the geometry in your output, but your expression will suffer decreased performance. Since you're grouping, you'd need to somehow perform a Union somewhere in your expression, and that kind of thing can get pretty "expensive" in terms of compute time.