Latest Contributions by jcarlson

Number selectors always operate in a way that ignores nulls. You could get around this with a Grouped Values selector looking at the same field. This might result in gaps in the numeric range for values not present in the data, and also means you can't use a nice "slider" selector, but it would preserve nulls in the data. ... View more

I like a good GroupBy statement in a Data Expression, as it evaluates a lot faster than feature by feature. We also want this information to be available for filtering other widgets in your dashboard, so we need to keep the objectid or globalid in the output. Inside of the GroupBy expression, we can also use some basic SQL. In this case, we use "CASE WHEN <condition> THEN <result> ELSE <fallback> END". It could be a long expression, but essentially, anything null counts as 1, non-nulls as 0, and the final value will be the number of null fields. Anything above 0 meets your criteria for "incomplete". The output of this expression will have one row for every feature in the layer, and you should be able to use it to filter other widgets accordingly. If you'd rather just have an expression that evaluates to "complete" or "incomplete", that's also possible with some adjustments to the SQL. var fs = FeatureSetByPortalItem( Portal('your portal url'), 'itemid of the layer', 0, // layer index ['objectid'], // since we are using GroupBy, we don't need any fields in the FS false ) var sql = ` CASE WHEN first_field IS NULL THEN 1 ELSE 0 END + CASE WHEN second_field IS NULL THEN 1 ELSE 0 END + CASE WHEN third_field IS NULL THEN 1 ELSE 0 END ` // group by objectid to keep every feature separate in the output return GroupBy( fs, 'objectid', { name: 'null_fields', expression: sql, statistic: 'SUM' } )   ... View more

For whatever reason, it's just not seeing attachments on the feature, if there are any. Are you using a View Layer? It's possible to have a view layer that doesn't expose the attachments of the source data. I would open the data table in your Portal to confirm that there are attachments on the features. I'd also take a look at the layer properties in Pro, just to make sure that Pro and Portal are referencing the same exact data source. If you're using views, or if you published the layer from Pro, it's possible that they're not looking at the same data. ... View more

And you can confirm that there are actually attachments on the given feature? Try adding the count of the attachments array to the output, just to check. When you create the array, put the count in as the first item: var link_arr = [`${Count(attachs)} attachments:`]  Do you see a count of items? Also, try putting some static text in the HTML. The attachments ought to all have names, but if the attname variable were empty, the link wouldn't render. var link = `<a href="${atturl}" target="_blank">File: ${attname}</a>` Any change? ... View more

Hard to say, but I think the href needs double quotes? The way Pro evaluates popups is very different from the web map sometimes. Here are a couple things to try: First, push your attachment URLs into an array, then use Concatenate to return a formatted block of text. Second, use template literals to build your HTML strings. This will avoid any issues with quotation marks. var layer = "https://somserver.com/server/rest/services/mylayer/myfeature_PT/MapServer/0/" var attachs = Attachments($feature) var foid = $feature.objectid var link_arr = [] for(var a in attachs) { var atturl = `${layer}${foid}/attachments/${attachs[a].id}` var attname = attachs[a].name var link = `<a href="${atturl}" target="_blank">${attname}</a>` Push(link_arr, link) } var links = Concatenate(link_arr, '<br>') return { type : 'text', text : links } ... View more

What's the context for the expression? Is it in an Arcade popup element, or just a standalone expression that shows up as a field? Is it the new map viewer, or the classic one? Any chance the data is public? ... View more

Unfortunately, your field formatting in Excel doesn't matter. When AGOL reads in your file, it does it all as text, then makes its best guess about the field types based on their contents. The best way to go about this, in my opinion, is to create an empty layer with the schema you want first, then use the contents of your file to append records to it. ... View more

Does the device / user switch happen at defined points? I think using multiple forms for the different sections could be a workable solution. ... View more

How do you mean? Would this be a single user switching devices, or multiple users dividing up the survey? There are lots of ways to collaborate and work on the same data in Survey123, but which route you go really depends on what you need the process to accomplish and how you want it to function. You could have the different sections as entirely separate forms, which you could edit through the field app's Inbox. ... View more

I don't know about a way to do it within the API, but you could just use requests to ping the service endpoints and parse the response. ... View more

It's a method on the GeoAccessor, Esri's version of a spatial dataframe. https://developers.arcgis.com/python/api-reference/arcgis.features.toc.html#geoaccessor You can also use pandas.DataFrame.compare() to compare two dataframes, but it operates differently. ... View more

Can you share the formatting you have on the HTML? Hard to diagnose what's going on without that. ... View more

Given that it would edit your features in place, it would work quite well. Some of our layers we update this way are used in an offline capacity, and there is no issue. When the source data has adds or deletes, those are made, everything else is simply edited in place or left alone, so any internal fields used to link features to the offline copy would still be in place. ... View more