Talk to Experts Tuesday - Qlik GeoAnalytics/GeoOperations FAQ

[Jamie_Gregory]

This is the FAQ for the Talk to Experts Tuesday Qlik GeoAnalytics/GeoOperations session on February 9th, 2021. The recording and transcript can be found here.

What is the roadmap for Qlik GeoAnalytics in Qlik Sense SaaS?

When we started out with GeoAnalytics, it all started out with the extensions, the connector and the server. These are the three things that make GeoAnalytics work. Since then, we integrated much of the mapping functionality or the visualization part inside of the native map chat. So, that is how we will take that farther. So the extensions will diminish over time.

As for the connector that extends Qlik Sense and QlikView’s operational abilities when it comes to spatial, they are being superseded with the GeoOperation service. So that is how we will do things in the future. I hope in the future also that we can see GeoOperations for QlikView and Qlik Sense but right now you are using the connector for QlikView, Qlik Sense on premise and the GeoOperations on SaaS.

What I’m hoping to do in the near future for GeoOperations in SaaS is to make it a bit simpler. We will hopefully allow for accessing GeoData directly on data file. So, you can drop a shape file or JSON file in your data files and that will be much easier to extract. I’m also hoping that we will have a small connector where you can add a URL or point to a file and then it produces the GeoOperation line for you so it will be easier to load Geo data. Of course, there are many things that you could do with GeoOperations but that’s the things that I can oversee right now.

There is also a Qlik Insider webinar coming up on June 3^rd for Qlik Data Analytics that might give you some more information on what’s to come. The registration page is not up yet but should be soon. You can find this on the Events & Webinars page on Qlik Community.

I have a question if it is possible to use Signed License Key for multiple QSE implementation which are geographically separated? Would be possible to get a contact a discuss it in more detail, please?

With GeoAnalytics, we are using the LEF key today, the text based key, in the server and the extensions. It’s important that you use the same key in the server as well as in the extensions. So those of you with the signed keys today, you can ask sales operations and get LEF key that you can use for GeoAnalytics. It’s a bit of a manual process to contact sales. For the coming release in May, we will be able to handle the signed key in all parts of GeoAnalytics but it still stays the same that you have the same signed key on the server as well as on the extension side if you are using that type of setup. If you have multiple setups, you shouldn’t blend the different signed keys, you should just use one signed key. That server can then server multiple Sense environments.

Can we compute area and length with the Qlik GeoAnalytics connector for Enterprise on Windows?

For area calculations, that type of calculation we don’t do in GeoOperations today. The closest to the calculation is the route calculation. That we do, but we don’t do any area calculations in Sense or GeoAnalytics and we don’t have any plans to add that I think. Area calculations is somewhat special, I haven’t heard so many asks for that among the client base. Of course area calculations is a well-known field in GIS but I’m not sure if that’s the number one functionality that we should add to GeoAnalytics today.

Please search/add your idea to Ideation so that interest can be tracked.

Complex intersections often fail and I sometimes have to resort to Python workflow to solve. This is for the connector on Enterprise on Windows.

I haven’t come across that much. It might be that the customer is trying to do intersect with really large polygons, maybe that’s the problem. I’m not sure whether he’s running cloud or his own server.  It might be the case that he has to increase the available memory for the server process (please see How do I configure available memory for Qlik GeoAnalytics Server). But intersect is a heavy operation. You should make some decisions rather is it necessary to have that high resolution on the that you calculate intersect on. It might be possible that you could simplify the areas a bit before you make intersect. It doesn’t work in all cases. Sometimes you need all the breakpoints. Especially if you blend several sources of information, the it’s good to have a common resolution for across all of your areas that make sense. So you’re not having one that is super detailed and bogs down your computations.

If you can provide an example, please post it to the Qlik GeoAnalytics Discussion forum and someone can help.

Where can we find documentation and examples on how Qlik GeoAnalytics works with QlikView? There is limited examples from what I can find online.

You can find all of the documentation for Qlik GeoAnalytics for QlikView on the Help site. You will also want to check out this post, New to GeoAnalytics? as there are a lot of references available here. Also, there are some QlikView Examples that provide QVW’s for you to take a look at. If you have any questions, please post to the Qlik GeoAnalytics Discussion forum and someone will be able to help.

I have geo data coming in from ESRI. And I believe that data comes in US feets. Can you show us how to convert it into Qlik readable format using geo connector?

Here is instructions on uploading an ESRI file that provides a sample ESRI file as well as a QVF and QVW.

When you’re talking the data coming in as US feets, normally you need more information about the coordinate system than that it’s just based on feets. There is something called the EPSG code which is a shorthand code for the reference system that you’re loading the data from. Sometimes when you load data, the information on the reference system is incorporated into the file but sometimes it is missing. So, when you run the connector, you will have to enter this code and that is sometimes the information is lost. Then you have to go back to the source and find out which is the reference system and there are thousands of reference systems to sometimes these can be hard to guess. In an ESRI shapefile there is usually a file called PRJ file, which holds the projection information. Sometimes it’s just variables and sometimes it’s names that you can figure out the code for. Normally you would do a search as well.

Maybe you get the coordinate system is not NAD83 and then you do a Google search to find the EPSG code. There are sites that will interpret the PRJ file and get you the code that is necessary. Given that you have a PRJ file, they are not necessary.

Where is the documentation for GeoAnalytics with Qlik Sense?

You can find all of the documentation for Qlik GeoAnalytics for Qlik Sense on the Help site. You will also want to check out this post, New to GeoAnalytics? as there are a lot of references available here. There are a lot of QVF files provided as examples that you can try to recreate as well.

Would it be possible to control zoom levels dynamically?

It’s not possible today to do it programmatically make the maps zoom in or out. It’s either you zoom it by user interaction or you have auto zoom on and then it will adapt the zoom level to the points that you have on the map. We have been discussing whether we should have some kind of API towards the map chart that you read or set certain variables. You could then figure out the zoom level or change the zoom level of the map. Nothing is decided so it might be possible in the future.

What you can do, you can decide which layers that should control the zooming. So, you could have an invisible layer that you are manipulating indirectly to make the map zoom in and out based on the bounds of this hidden or transparent layer.

My license says I have GeoPlus, do I need to install it? What is it for?

If you have a GeoAnalytics plus tag in your license, you should install it. Here are the instructions to install GeoAnalytics Plus. You must have Java installed and it must be 64-bit. It also must be at least Java 8. That’s very important because GeoAnalytics depends on Java for communicating with the server. Also, make sure you only have one version of Java installed.

You can find the benefits of GeoAnalytics Plus on the Help site. Primarily, you can connect to local databases and limits are removed for process with this connector.

What resources are available for GeoOperations?

We have a whole website about GeoOperations. GeoOperations is done on the backend in the script. It is especially focused on our SaaS offer. We have a whole bunch on how to use any operation, how to migrate. There are people that are used to using the wizard but now have to translate operations in this script so it shows how to convert that. We have more examples in the Community as well. If you have any questions or need assistance, please post to the Qlik GeoAnalytics Discussion forum and someone will be able to help.

How do I Geocode?

For Qlik Sense on Windows you can enter addresses into an inline table or you can bring addresses from an excel spreadsheet, a database or csv files. You can then use the GeoAnalytics connector and choose AddressPointLookup and use the wizard. Match the fields to the fields in your table. Then click Next. Then insert the script and run the app. You can then plot the point. To see a demo of this, please watch the recording at 30:45.

You should also use our template because we want you to save your results. It allows you to store the geocoding results in between sessions so you don’t have to reload all the addresses every time. You only need to do it once and then it stores it into QVD files.

So with this template, first I load the addresses then I try to load the cache if there is one, otherwise it will select the new ones around the lookup operation. You’ll get a quality match on how good the input address matches the output address. You will also get the matching level whether you get it on house number or just on streets, you get to match on the street or if you get it on the zip code or on the region or on the country, which is the lowest form of precision. To see a demo of this, please watch the recording at 34:30.

Some tips you should consider:

• Make sure that you have data fields with proper information
• The quality of the lookup is highly dependent on the data
• Make sure the address field only contains the address, not phone numbers or apartment numbers
• If you know for example, the State information is not good, then you should omit that information rather than put it in
• Sometimes abbreviations are used for streets, you should try to write that out if possible
• Zip code or postal code has really high precedence so if you suspect that something is wrong with that zip code information then you should leave it out
• If you have a large amount of data to geocode, start with a smaller subset to make sure everything loads properly and all my field names are correct and the data looks fine. Then run and look at the results.
• Between 5 and 10% of addresses will be faulty. These will be filtered out and you can manually fix them (preferably within the source)

Address to point and point to address is not part of the standard analytics so it’s a subscription that you acquire. It allows you to do 1 million lookups during one year but you are allowed to store the results and use them for other purposes. For more information, please contact your Account Manager.

Can the UI "get" the current zoom level?

That ties into the previous question regarding how to control the map chart. You can’t do that today. If you have auto zoom on it will automatically zoom into the bound of the current layer. If you know which layer controls the map, you could calculate the bounds of the points and from there calculate backwards the zoom level. The actual zoom level is hidden where you have to switch it over in the Properties panel and you can switch on the debug info. Then you will see the current resolution and zoom level where you are. Normally you would use that information to control which layer shows when you have switch layers on and off by zoom level.

Can you set background layer (ie satellite, pale, default) dynamically?

There is this exploration menu, here you can control the different type of base map layers. That is one way to do it. Another way to do it is to programmatically control and you can switch between the options.

A basemap layer is where you set the endpoint with a URL. In this case I am using a field to control the URL but you could use a variable instead if you want. Here is an example: Setting up a TMS as a background map in Qlik GeoAnalytics.

Is Auto-Select visible available in "native map"?

So that is a feature that we had in GeoAnalytics for Qlik Sense which allows you to do selections based by zooming and panning in the map. It uses a spatial index. So that is a functionality that we haven’t brought into the map chart. In the future I think we will do something similar to how the scatter plot works. With the smart data compression, there are engine functions to create binning layers and load data on demand. So, we will probably use that instead. But it’s not available for QlikView and it’s not available in the native map chart yet.

Is there an end-of-life for the GeoAnalytic extensions?

We have not set up the end-of-life date yet for the GeoAnalytics Extensions but it will come. I encourage everyone to use the native map chart instead, at least for Sense because in QlikView they will stay. But for Sense, the native maps now have much more functionality. We have the WMS wizard, we have background layers, there is a lot more functionality. Printing works in a different way. Export is better as well. So I encourage everyone to use the native map chart

Are lat/longs required to plot data?

I am assuming that question is about using the native map. No. You can use any named area or named point. For example, if you load your city name and map it, it will show your city. Maybe you have to adjust your country because for example we have one state here in Brazil that is AL for us and you have a state that is AL as well. So, you have to set the country properly.

What is the difference between GeoAnalytics and GeoOperations?

GeoAnalytics is a platform. We are talking about the extensions and the connector and the operations. But as Patric said, as we move forward it will be about the native map chart and GeoOperations service and the analytics server. There are many advantages to using the GeoOperation compared to the connector.  The connector is based on the connector API which is not really meant to do the type of operations that we do. For instance, it’s much easier to load a shapefile. It’s much more condensed code. When you do an operation on a loaded table, much of the code that the connector produces is for chopping up and sending the parameters to the map server and back. But now we are using the service side extension API to stream the internal table to the map server and result back. It will simplify a lot and it’s faster and more efficient.

Uber's H3 Hexagonal coverage is stable (size-wise) and is available at many zoom levels. Might be an interesting supplement to current binning. Thoughts?

There’s many ways to do binning. We provide two different binning methods today. We have a square based binning that is degree based, and then we have hexagon binning, also degree based. But there are ways to do binning. You can actually create your own binning as well by making computations. If you produce binning layers in a different environment, you could also incorporate them into your analytics. I don’t have a specific option on Uber’s product but I will look into it and see if we can make an integration.