Q&A with Qlik: App Development - June

[Troy_Raney]

Environment

• Qlik Cloud
• Qlik Sense Enterprise on Windows

Transcript

Hello everyone and welcome to Q&A with Qlik. Today we're gonna be talking about App Development and all the things that are involved with that.
My name is Troy Raney. I work to help put together videos and webinars like this one and we've got an excellent expert panel for you today. Ola, why don't you tell us a little bit about yourself?
Yeah, sure. Thank you, Troy. So, my name is Ola and I've been with, Qlik, for a little bit over 7 years in various different learning teaching roles so I've been a technical trainer and now I'm working on creating online content mostly for Qlik Cloud but I also dabble in some of the older products.
Great and Patrik.
Patrik Lundblad, I work in Product Design.
I don't really do the UI design or UX. I'm more there working with the design team and architects around how do our users use our product; what's the value you want to get out of it? Just explaining things to people around what's the usage is. And I've been here for 11 years now.
Great. Alright, first question: I understand how to use the MIN and MAX functions, but how do I specify data that is between the first and last, like the second or third item?
If this data is a text string and has no numerical value, I'm not sure how to do this. Thanks.
Well, there's a couple of, even more details that would be wanted to be known in this case, but I, if you look generally, if we're inside of the, the data model, in the data load editor and working there, and we want to find specific values or rows within a table.
We could use functions like a row number and peak. To be able to find specific positions within a table.
And that kind of in that case we need to know that they're sorted in the way that we want to have them sorted if it's a specific.
Value we're looking for in, in an order in this case in a, in an order of stream values.
Okay.
That's great. Thanks. Okay, next question: How can I find the second top 3 values in a field? And there's a example of a dataset here with dates. Product values associated to each of those. Any ideas on what would be a good option for that? And I'll just, thanks.
So I can take that. So, I've talked about the RANK function and the role number. Another function we have is the RANK.
So taking a dimension and a measure you can have a rank you see which is the rank 1, 2, 3, 4 and so on, so did a fun example if you stop sharing I can start sharing my screen
Yeah.
Where I used the rank function to identify if a value within a certain rank position. In this case here, I have a chart with some Sales for some customers, of course all training data or dummy data. And I decided to use the rank function for coloring.
So in this case, I'm saying if the rank or sum of sales Is larger than in this case I'm using a slider but if it's larger than 0, and smaller than 4, then I want it to be one color compared to other colors.
So rank function together, we have statements, get, and function in there. And then I used it with a variable in here to be able to highlight which is the top 3, top 4-6, and so on.
To rank functions, extent of saying if you're not top 3, if you're a second top 3, and I think that's what you wanted in this case.
Excellent. Thank you. Okay. Next question:
I want to create a variable that can help me select specific weeks. For instance, if I select 202402, I want to retrieve the past three weeks: 202401, 202352, and 202351. However, my current approach is giving incorrect results. Any ideas?
Yeah, and so in this case it's also a little bit dependent on where those how those values are created if they're read as dates or not or if they're just a read as a concatenated text strings, but if we have access to a date field within the data load or in the data model but we have not generated feels in the correct way we can do that in inside of the chart functions as well so we can create our own week combinations of week and months and years, if they're not working in the way that we want from the data model.
So that's possible and then after that we can add, subtract or, 1, 2, 3, 4, 5 from those values to, move back in time, so to speak. I would also generally want to go back to the data. Model. And maybe even if I don't have access to it, maybe go back to it and see if I can create flag fields within the master calendar if there is one, to be able to make these calculations a bit easier
Great, thanks. And Next question: Is there a way in Qlik Sense or Qlik Cloud to load sheet IDs or sheet titles similar to QlikView using document summary loads?
I would say, maybe Ola know. I haven't used QlikView that much, so
I would say, not, not, today. I'm a little bit unsure.
I think I'd want to send that to community in this in this time.
Yeah, speaking of that, there's some a great resource. I don't know if you guys have been there, but on Qlik Community.
Community.Qlik.com. There are forums up here. Especially analytics and they're all very specific topics but there's one on App Development and if we have a question that we're unable to address here, you think of right after the session ends. This is a great resource. There's over 200,000 users that are active on Qlik community, so It's always good place to post questions and get answers.
Alright, next question: Can you describe the best practices when using the subfield function in the data load? I'm trying to get count of items in a string field with comma separated string items.
Sure. That, I've always, and I mean, I've, taught the Qlik, scripting and I've always felt just that, a syntax being a little bit tricky.
Getting the commas and the, everything in the right place. I would refer to looking at some of the examples of that.
That we have, on Qlik help. To kind of understand and break it down. On how to use the limiters.
I think there's some really good examples up there. To kind of look at a little bit on how the syntax works and then if you're having further problems with it or if there's something that you don't understand that and that they're going to community.
But I think that explains it in a good way.
Okay. Next question: Does the CSS styles and multi KPI still work and Qlik Cloud?
I believe it should so. I don't see a reason why should not at this moment that I know of.
The main thing to think about when you're overriding the CSS of the product, the UI is that we sometimes update the UI going from Angular to React, adding new components, renaming components and so on.
So there's no guarantee that it's gonna work tomorrow, mainly because the CSS names that you're using in your custom CSS might get a name change or something like that.
Yep.
Okay, thanks. And Next question: how can I show percentage of total in a bar chart?
Yeah, so I did a small example now. Let's see if. I got the question right for first of all.
So I have a bar here. I have some customer segments. I have a percentage here and I have sum of sales.
So what I did in this, simple example is that I just took the sum of sales, which would be for this dimensional value.
So it's doing sum of sales over each customer segment. And then I divided it by the sum and then including the total in here.
We cause total I'm gonna do is gonna aggregate all those values together. Not only the road that it is on in the data cube or a hyper cube at that moment.
So taking sum of sales to my device, some total sales gives you for each bar the percentage of total.
And you can use that calculation also for tool tips or for another place where you wanna have a number displayed to the user.
Hopefully that's also the question.
Yeah, that's great. Thank you. Okay, and next question: What about backup of apps with or without data and Qlik Cloud? Visualizations, the sheets, the master elements, variables, etc.? That always comes up.
Yeah. Yeah, exactly. I think, yeah, that's a really interesting topic as well.
We have something called automations that you can work with in, Qlik Cloud and what that allows you to do is to connect to external sites like a GitHub, for example, which you can use through connecting to the API you can extract or you can export your application without data but with all master items and all visualization, all the data about the visualizations and then you can import them there later.
You can set up that as an automatic, and then you can import them there later. You can set up that as an automatic, task as well.
So for example, every time you reload the app or, make changes, it can fire this automation.
We actually have some great videos of that on, Qlik learning as well on how to set that up.
Great. Next question: About linking tables. Here's the first scenario which is easy.
I have user tables, location tables. And accident table which has the location ID and user ID. This is straightforward linking, no issues, but in the second scenario. There's an accident stable. It's been created by and a modified by, in this scenario. How many claims an accident table can be linked to users table not using the apply map. It's hard to visualize all those details, but…
So yeah, but I think I understand kind of the basic. Problem here with the question and then when it comes to Qlik data modeling really basically speaking and then this is again what separates Qlik in some, some places from other, other data.
Tools and it's that we automatically try to connect tables to each other. By, renaming fields to have the same name.
Now, a Qlik doesn't automatically check that they have the same values, but it checks for the same name.
When we come to, more complex data models where, we don't have necessarily have one key field in a table, but it's, rather a, field in a table, but it's, rather a concatenated, field.
We need to. Find those concatenations in both of the tables. So we need to find the values that co-exist between both of the tables and then rename the fields to the to have the same name and then Qlik the fields to have the same name and then Qlik automatically will connect them to each other.
But it's hard to link the table. With another table if you don't have any fields in common between those tables and that in that case you need to go in and do a link table or some more advanced syntax to get that up and running.
Okay, great. Next question: If I have dimensions field A and field B. And a measure score, how can I get the top 3 scores per field A and per field B?
I would say it's still the rank function. I don't know if it's that feeling of you'll be our dependent on the shutter also if you need to do it in some special way they're off combining them together or so on.
Yeah
I think.
I usually, usually when someone says top something, I always say RANK. There are probably other ways also.
Okay.
I think it's a great place to post some example data and get me or other people to respond to the expression needed for it.
Yeah, I think that if we're looking at an already created table, like a bar chart or something like that, there's built in functionality to get the top.
X values from the dimensions there as well. So that's just to expand the dimension and you'll find where you can go in there and limit the amount of dimensions that I shown and then for example take the top dimensions or such.
Hmm.
Great. Thanks. The next question: is there a difference between a script variable and a Document variable.
And Qlik Sense.
There is a difference between the variables that you create in the script and the variables, the variables that you create in the front.
So usually you use variables when you script, to, do what you do with where temporary values and such, forth.
And some of these value, variables can then be accessed through the front end. the, values that are in that variable is changed or updated every time you run the script.
So, every time you rerun the script, it'll have the values.
That it goes through and does the logic in the script. If you want to separate. Variables that you use when you create the script with variables that you use in the front end.
I recommend looking at a naming convention in our case. I do believe we have some examples of naming conventions in our QVF framework.
Qlik deployment framework that's available both in a QlikView, Qlik Sense, and Qlik Cloud version. So you could look at that as something for some examples or, yeah, on how you can work with naming variables differently in the script than in the front.
Great, thanks. Next question: is it possible to share or cooperate on an automation?
Yes, it is and the reason why I'm smiling is because this is something that's been requested for a long time but just recently was sent out the gate so yeah it is
That's fantastic. So a new feature available, awesome. And Next question: is there an option to live query data? For example, a schema in snowflake?
So if you talk about live praying, seeing consuming data live, there's when you create and Qlik Sense apps, there's generally 2 modes of data.
One would extract the data, put it into the in memory, and then it's part of the app as a QVF for QVD and so on.
There is another mode also for some of our connectors to Snowflake, etc, where we have a direct various that.
Where when you're creating a new app, we don't actually, we don't enable everything that happen, all we have on apps because everything is not supported because it's not in a memory.
But then you can use the live connections. That each has a live connection back to databases. I think we're calling it direct.
Query or something similar like that that's available in the help pages. And I can try to find it and we can link it in the webinar afterwards.
Awesome. Next question: I have many apps in the server, but one of the apps is taking a longer time to load. What are the measures needed to be taken? I'm storing the large tables into QVDs. Any suggestions to help improve Reload performance?
Yeah, I mean they're here there's a lot of things to think about when looking and their articles written on this, the, the, you know, Qlik deployment framework, for example, is also something that you can look at.
But generally speaking, if you want to check. Some reasoning behind why it takes so time, long time to load. There are monitoring applications available from Flick you and Qlik Sense and the cloud as well. Not all of them are installed out of the box, but they're all accessible. In one or another way. So I would recommend start looking at some of those monitoring. Going in looking at the specific app and then seeing what parts of the app is taking, such long, long time to load and kind of start there.
And also something we released quite recently. It was on cloud last month. Is, pre warming, we released quite recently, it was, cloud last month is, a pre warming where you can actually after a reload, you can have multiple of the calculations being calculated before I use actually goes into an app.
Think of it as an agent that goes in or a both that goes in and does some of those calculations that they will be cash for.
I think we have similar solutions for size managed either built in or by partners. Or thank you, that's why you don't have to manage over long.
That we launched it. What's it the May release or February? So there are ways of pre-calculating things also set up and app those quicker for the first user in the morning
Very cool. Thank you. Next question: Bookmarks are very crucial to the customer for repeated Selections. How do I ensure I don't lose them when I make changes to the application? Or export the application from on-premise to Qlik Cloud.
It's a great question. I do know that we have a Techspert Talk session later this month. Where it's all about migrating from Client-Managed Qlik Sense Enterprise on Windows to Qlik Cloud and we're going to be talking about migrating, publishing and migrating apps with those all those personal items. Private sheets, bookmarks. Storytelling and how to get all those to Qlik Cloud in the same application. Sometimes it can be. A lot, but, did you guys Have any suggestions?
I'm being, you can always, save these items. I mean, they, they are available even if it doesn't seem like they are available, when you just look at what an app is, the all the personal items are also accessible and in different ways.
But, I would, as I, if you're looking at making a migration for first off, attend the webinar or watch the material because there are tools in place to solve this for you.
If you were to kind of just try to recreate everything on your own. I also recommend the process of going out and talking to your users and giving them a time limit for where they need to kind of save their items and Look at what puts their individual stuff that they really want to want to keep.
But I mean, it's a complicated question.
Okay.
Yeah, just can be a complicated thing, but It's totally possible. And yeah, if you want to see that demonstrated, please tune in to the next text per talk.
Next question: I have a transaction file with many fields, but the key ones are a value field. Field holding a CPT code and a field holding the authorization number.
I want to calculate the average for each authorization grouped by CPT code. So each CPT code will calculate the average value of each aggregated authorization number.
Wow. The part I do not know how to do. That I want to eliminate the top. And bottom 5% from being included. Okay.
So I would say give us some data examples, put on the community, me or other people can help you there.
And the one thing is that we have a function called Fractile that you can use to see if something is like fractal(), the measure, and then 0.05. Would show what is the limit of the 5%. You could get that quickly by fractal.
Thank you, use, if statements to see if it's the number is below or above that fractal value and then you can include it or exclude it in your calculation.
But these are things that are easy to do having a data example up and typing it.
Yeah. Course. Okay, a next question: What are the visualization recommendations for a dashboard design for analyzing survey responses? Integrated with findings from machine learning. Any suggestions for apps? Analyzing survey responses.
Very quick question, I would say. Depends on survey depends on the amount of responses you have things like that simple service like people put a scale of something, a bar chart with a middle line that says if it's positive negative.
Usually used if you wanna have machine learning together with it. Table on the side, I can explain things.
I would say, yeah, depends on what data you have. What do you wanna show to the user?
Maybe not saying what data goes with what chart rather. What does he just wanna find out of the data and then seeing what chart is applicable to it?
Yeah, that's a great answer. Okay, next question: How can you show a model filtered by a login user? And a PDF file produced by a report filtered by the email address that it sent to. Okay.
And there's a follow up in the next question also.
Thank you. This has to do with people being able to see only the information they allowed to in a particular model. Oh, sounds like we're on to section access.
All right. Yes, there we have something called section access, which is, something that you set up in the script of your application where you can set up a table limiting a row and column access to information without throughout the application.
So you can limit the possibility for a user to see specific values. In a column or to take away the actual column itself.
From the information that they can see. And we then use section access throughout other functions throughout the platform as well.
So if you, if there is section access limiting the amount of information that you can look at in an application, there's no way to kind of circumvent that and search for the information in another application or another part of the platform as well.
So it's something that then that I used throughout the entire platform.
Great, and we have a lot of great resources. When it comes to applying section access. I'll see if I can link something. To the recording here. Next question: is there a way to use actions when I leave as sheets not only on activation?
Not that I can think of, no. No Qlik you support that. We don't have that in Qlik Sense so far.
So that's also a good place to talk, about Qlik Ideation. Qlik Ideation is something we have at the community.
Yeah.
Just quick enough, you can navigate there. Where you can suggest your ideas to us and you can also see what other people have suggested.
So before you type in your idea, make sure to go to and see if anyone else has put an idea before.
And then you can read the ideas that are available in there. You can like the idea, then it will be plus one in the score for it.
And you can also see comments if people have similar issues or at the suggestions and so on. And I'm working with a design of course, but we have prog manage, work manager, we have other people in company who are watching Ideation, looking at ideas, looking at the top ideas and either getting feedback, putting it on a roadmap or explaining why it's something that we can't do.
So, great place to see if there is already the. On sheet leave. Action in there. If not, submit your idea and start getting your friends and colleagues to vote on it.
Yeah, definitely. This is a great resource. And when our product management team definitely weighs this heavily when they're considering improvement.
So Please provide your feedback. Next question: is it possible to load SPSS files? (.SAV to be more precise) to be able to extract the labeled dataset coded data sets and also the metadata of the existing variables from the .SAV files. I think up a now work for marketing research. It's considering abandoning Qlik because it has not been possible to do this without the help of scripts in .Net and or Python. Sounds like it's pretty specific requirement. But I'm not familiar with that. You guys?
The only thing that I know about it is that if there is a file format that's not supported by Qlik.
The in through the data load editor, they're mainly 2 ways to look at it. The first thing is to see if we have a connector that's able to connect to the source system and extract the data.
Right.
From that, the other way, the other thing is to, convert. The file and I think that's probably what they're talking about here, but convert it to like a CSV file or something like that as well.
And then I don't know if that's what they're referring to with Python and Dotnet language but otherwise.
Convert to like a CSC file or XML file or another file format they can read or connect to the system itself.
That's good. Suggestion. It sounds like that similar to the workaround I found, but We'll move on.
Next question: How do I get data to load in a sorted manner so that the row numbers can be referenced properly?
I like it.
Yeah, you can sort them in in the in the script. You can do that either. When you have been the data source before you load it, you can do it inside of the Qlik script as well.
By, sorting a grouping biases, certain values. It depends if it's on numeric field or a text field.
So from numeric field, it's easier. If it's a text field, it's a little bit more complicated, but you can use do that through the script
Great. And Next question: and I was just about to read, I saw that it snuck into the chat and I see the same question now in Q&A. Thank you. Is there a way to export pivot tables from formatting using Qlik Sense installed locally?
I want to export the pivot table exactly as it is being displayed. Okay.
So currently no, and I don't think there's any plan for the current main pivot table as far as I know.
We are working on a new pivot table that is in the visualization panel today that we are working for export.
I don't know. You sure it's gonna be a limitation between Cloud or on, and so on.
But I would say that's probably more where we're gonna make sure we're gonna move improvements when it comes to printing.
Okay, great.
But I can't say when or if, cause not my ever really.
Sure. But it's nice to know that this is something that we're concerned with.
Next question: How to show the totals in the bottom instead of the top in a pivot table.
So that is actually the same answer as the previous one. So if I share my screen
Yeah.
So we are working on the new pivot table. So if I go to. My example here. If you go to custom objects, if you go to Qlik visualization bundle, you have the new pivot table in here. And if you in the field table turn on totals, then within presentation you have show totals above or total those below option. So that's available for new pivot table and there's many more things to come. Example, we have the new value that many people are asking for, being able to say, is it a dash where it's a zero or something like that?
You pivot table also has a lot of specific stylings on the header, the dimension values, measure values, labels, total values, grids.
Yeah, there's a lot of options in there with a new pivot table and we are also looking and working on having a, the exploration panel.
We're an end user can include or exclude the dimensions, measures in the pivot table and do live pivoting because life is supported in the main field table.
We want to have the same thing in the new one, but that's going to be made from the Experation panel is that.
That's awesome.
Yeah, that one has it and a lot more. Maybe not everything but a lot more.
Great. Thank you. And next.
And it's available also on client manage, should be any different between client management cloud except for release days.
Okay. Next question: Is there any way to check if an app is using any column included in its data model? like with the tool called QSDA?
Yeah. I'm a little bit unsure. I would say monitoring applications in this case. You can see some of the data inside of these monitoring applications on which sheets are accessed also which, visualizations are accessed as well.
So I think that's kind of at that level. I am not familiar if there's any other way to to solve that.
Okay.
What it comes to cloud, you're working a lot on Lineage. So being able to see not only within an app where the data comes from but also if you have your data sets in the hub, you need to see where they're being used.
So that's an area we're moving. Quickly in, but that's on cloud only.
Yeah.
Okay.
Next question: What is the recommended approach to apply a single application for multiple clients with separate data sets.
Sounds like it might be section access again if you're Trying to have the same app with different data sets.
True.
Yeah, well it This is this is dependent on which platform we're talking about. I think, I mean, if we're in the cloud, you could use application automation to create local variants of the application like we did with loop and reduce for QlikView back in the day.
For example, otherwise either section access or actually having different instances of that application if the differences are too large or you can find that you need to make differences to the like the visual feel of applications and such as well it would be to make differences to the like the visual feel of applications and such as well, like the visual feel of applications and such as well, like the visual feel of applications and such as well.
It would be to maintain multiple copies or help the application.
Great, thanks. Next question: When I load a data model from the Jira API, I have a table with a string field.
Hmm.
Further, if I create a filter panel in the report on the string field and select any object from it, then the report will break and I'll be able to open it after some time.
You can also create a table load in line with the string fields and the same thing will happen. This a bug?
Does sound like a bug? I'm gonna test it afterwards But I would say do file a bug on that.
So it gets prioritized and looked at.
Okay, next question: How can a bookmark be deleted once it is obsolete and the user no longer works for the company.
If we're talking about on-prem, you can find those objects inside of the, hub, I think.
No, sorry, the, the QM (QMC). I lost the name of it.
The management, tool, for the server and if it's on the cloud the app owner should be able to see maybe not if it's personally created, but otherwise should be able to right Qlik and delete the most of the items there.
Great. Alright, next question: How can we protect our Qlik Sense script source code? Can we use section access to the script or is there any other method?
So I know that in QlikView, we used to have a hidden script. Where you needed to type a password to access the script and that's usually where you put the section access as well.
We don't have that functionality. In in in the in-Qlik Sense.
But I know at least in Qlik Cloud the manage space is if you publish something to a managed space there is no script right?
Yeah, and the same goes for if you publish an application inside of, the Qlik Sense hub as well and you're, you're not an admin of the app.
You, you, you can't access the, the script. You need to publish the app and get access in that way.
So in that case, you can say that the script is hidden, but there's no way to hide the script from other developers.
So if you have multiple developers or has access to going into the script, you can't hide parts of the script.
Okay, next question: How to share objects visualizations, sheets, etc, created from a professional user. Other defined users or other groups like object sharing and QlikView.
So we talk on Cloud, we have the right-click menu that has a share option and you can share link to do association and send it to someone else.
If we're talking clients or on-prem, then I think you can always go into the developer hub and you can find everything there the link to the object and you can send that to someone else. Of course, to be able to consume that vizualization, no matter if it's on
On-prem or on cloud, you need to have a license that You have some access to see it. Otherwise, if you don't have a license, then yeah, you're not allowed to connect to the.
And. To the app.
Great. One question I wanna make sure we get to. I think we did. Alright, next question: Is there an existing tool to quickly generate an applications design?
If we're talking about the cloud product, you could say that Insight Advisor has that functionality.
So with the help of inside advisor, you can get suggestions and charts and with measures and dimensions that you can use throughout the application.
And Yeah, you can just right-click and add them to sheets as well. So, I'd look at Insight Advisor.
It's great. Good idea. Next question: how to set up field labeling?
Yeah, so this was one of those questions we got before, so I prepared for it. And I think, the kind of me would feel labeling, but to me it often is about As a consumer, a user of an app, I do a selection in a list box or a chart and I get something odd showing up in the selection bar.
So I'm going to show you a little bit about how it works today, what other different scenarios and so on and also talk a little bit about what we're working on that will be coming.
So if I go to my sandbox in here. Just been my smart things. If you go to your master, you can see that I have dimensions here that are interest.
One is year and country. You can see that I have dimensions here that are interest. One is year and country.
You can see here that there's fields that I've concatenated together. Using the and symbol.
And then you have the dash in between. And there's another dimension in here that contains a pure field, so it's nothing that's There's not that feel on the fly, on the fly.
This is something that goes directly to the data model. And difference now is that if I'm using list box here, if I do a selection in your dimension, You can see that it is here that shows up here.
It is not the year dimension label that I have in the most. So in this case, it goes directly against the field and it shows the field name.
In the case that I have something that doesn't exist in the script where I multi multiple fields together and those fields are within the same table.
a field on fly. Let's see if I still have that. Yeah, just a blog post here about what is that cafe, compared to field or flight or I post also But in case that it is a feel on the fly, then the end you will actually look at the list of most dimensions.
And it will use that in the selection bar instead. So that's a way of naming fields.
Now of course you can cheat around this or do something like that. So let's say I have a year dimension in here if I open it up to edit it If I open the field in here, I can add something to this field that makes it become a feel on the fly.
This case I'm just gonna add in Empty afterwards. So concatenating the field with an empty string.
Then I'm saving it. Going back out here.
Yeah, now see, your dimension shows up instead because it's doing a selection against that field the fly instead of the field.
No, everyone in the chat who knows about performance and so on is gonna complain. Because I'm suddenly creating an extra field and I'm just and the data model.
So I wouldn't say that this is a great thing. To do. Is that is something that if needed if you can't rename the field and the load script then you can create a few on from it.
The good part is we are actually currently working on making this better so that if you have a field inside or if you have a calculator field inside the dimension list, then we will use that as a label up here.
And we are also looking at supporting an expression in here. So if you're using different languages within your app or something like that and you put a label expression on the master, then it should show up there also as the evaluated string instead.
So hopefully soon, can't say when, but you're probably gonna see it in cloud first and then on client managed.
That's awesome. Thank you for that.
Next Question: What chart object do you recommend if I want to show some KPI like traffic lights? Red is something wrong, green, everything is fine. If I only want to see the color.
It's a tricky question. Multiple ways of doing it. I would say the simple thing would be for me to have a table and use.
Inside the measure you would use or dimension, you would use the symbol, a circle. Just find the ASCII.
Just find the ASCII color or asking symbol for circle, put it in there as a measure, then do a call, put it in there as a measure, then do a color by expression and have that one colored. Then do a color by expression and have that one colored.
Maybe if you put 1 next to each other and have that one colored. Maybe if you put 2 next to each other and color depending on, yeah.
Things like that. But I would say Yeah, some chart that uses caller by expression to do it.
There you go.
Or you can do, a tab container and having tabs and then having of those hidden and one showing and you have a picture in there or something like that.
I don't know how many KPIs you have actually wanna do it for.
Okay.
They're all good tips. A next question: With Qlik Sense on Enterprise if we want to export data and send it via email or so on, we use NPrinting. What's the equivalent in Qlik Cloud? There's definitely some options, but go ahead.
Yeah, yeah, but we have Qlik Reporting, which is a tool that's both built into the platform itself with the possibility to write Qlik on visualizations and setup that on an application level.
Together with Application Automation. So application automation Is an enabler for multiple tools that we have.
We talked about it before for exporting applications to get up, for example. In this case, you can use it for schedule and reports or sending reports out.
And there you have the option and opportunity to sending reports out and there you have the option and opportunity to use a rich, there you have the option and opportunity to use a rich, API, which the option and opportunity to use, rich, API, which amount of API is to do that where you can export API, which amount of API is to do that where you can export it as different file types, but also different systems
as well, like teams or Send them by email and stuff like that. So, Qlik reporting is what, what we call and printing in the, in the cloud and Qlik reporting is a combination of, functionality that's inside of the platform and application.
Automation to make it all. Spin up.
Yeah, it's very powerful. And there's A lot of different ways to send charts or Yeah.
Definitely check it out. And I see the Next question: Patrik, you're already starting to type but it looked like it was made for you as someone's asking about.
How to set formulas in their bookmarks. To show now for a date so they're a bit more dynamic.
Yeah.
Actually, that's of my favorite press fields. It's how it comes up. I think I did it in the last day one or if not the one before.
Yeah.
I it also internally because there's many people who don't know. What a book more case and how we save selections.
So I go back and share my app again. Going back to the same app. Yeah, actually gonna see here that this entire app is actually called.
Tips and fix bookmarks because I created to Talk about, this one. So let's go to the book more tab.
Of course I have an incomplete Yeah, and remove some dimensions before. Let me ask quickly fix that by just doing a Let's do a list box.
Let's do year.
Let's share a selection. So I want to select the latest year in here. I want to save it as a bookmark.
What happens when I do, select the value? If I create bookmark up here and I say,
Latest year selected.
Then if I press there, okay.
You can see her that there's, some information about the bookmarks like the, say, P, you can also see here there's a second question that says year equals .
So this is what I call a static book part because it has a value that's included in the bookmark.
Rather than a expression, something like that. If you want to have an expression included in it, you need to do it in a specific way.
If I share the selection in here, if I open up the search icon here. Hi, and then it's the field name that I want to do the selection in.
And then I tab equals and I'll type max parentheses. Total year?
So I'm doing now is that I'm using a search expression to say for the field year, find the max value out of this.
And what's important now is that I don't actually Qlik the value and selected. Once I'm down here, I Qlik enter key to have it selected.
And then if I go back up to my bookmarks, create new bookmark. Then we can say search expression as the name of this bookmark instead.
If I create it. And if I close this one here. You can see here that it is actually an expression that is part of the bookmark instead of the field value.
So that's how you can create a dynamic bookmark. Have it applied as a default bookmark that selects the latest year or something like that.
So search your expression is the way to go. And if you have already selections in the field, things like that, you can have to invoke only bracket one and things like that instead.
But the main difference between Search expression, static. Don't press the value. Type it in, press enter.
That's great and very powerful and So simple once you know but Yeah. Next question: Is it possible to create a pivot table with multiple dimensions which displays the dimension value.
And all rows of each dimension rather than grouped.
There's a follow up to that to question us down as well explaining it. Yeah.
Thank you. So essentially, as dimensions appear in a straight table, but with the column benefit of a pivot table.
Okay.
Okay.
I, I mean, if, I would go to the script, but that's, that's how I'm raised and, and, all those dimension values inside of a, field and, and then use that as kind of a, all those dimension values inside of a field and then use that as kind of a data island within the data model, but that I can use I don't
know if you have any way to solve it in the front end, Patrik.
There you go.
Yeah.
You can always concatenate fields together like I deal with a year and country. I used to feel I used the and the sign and then I used another field preferably having some type something in between there that, sign and I use another field, preferably having some type, something in between there that space or whatever that of course I think that's worse performance and doing it in loads quick.
Anything you do in little script is always worth doing from a performance perspective. I'm sorry to do it in the front end.
That's a great tip though. Most people aren't really aware of that.
Yeah, I guess that the argument for doing it in the front end just to take that general, bounce, would be if you need the real time aggregations based on the dimension values that you select.
So you want it to be updated because you give the static response in the in the data load editor so you can generate the fields out of the entire data model but if you want to have it more dynamic you'd want to do it in the in the front end.
Great. Okay, next question: Your CSS naming, for example, in the visualization bundle doesn't seem to follow a structure as opposed to the integrated charts.
Are those temporary like the new pivot table that's said to be included in this? With the standard charts.
So I guess asking about the naming convention with the bundles.
I wouldn't know actually. I think that's more helpful. Give to developers. There are different development teams working on them.
Some of the vulnerable associations are actually made by community members and then Adopted by Qlik and introduced into the product so The bundle is a mixture of Thanks, that other people created that we were really lots of usage and we decided to support it for you to things that we are being developing like the new pivot table.
We're putting in the association bundle because it's not fully there yet to replace the current pivot table, but once a chart becomes good enough, we move it from the bundle into the native charts and then maybe we'll do some updates on the in there.
Like I say that the bottom of it we have today is a great example. It was made by a community member, or actually ex, employee.
To begin with, we Structured the code, put it into the bundle, and then we did an actual native version of it.
So, yep. So that for Sometimes things we're working on sometimes things we have adopted.
That's interesting to get that insight. Next question: I'm trying to use partial load and Qlik Cloud app. Which contains calendar derived fields. Everything works fine on a full reload, but when doing a partial reload, the calendar fields are lost.
The derived field are lost and the app becomes unusual. Do you have a solution for this scenario?
Yeah.
Maybe a little difficult without seeing the script. What exactly is being partially reloaded, which part that kind of thing.
Yeah, exactly. I think, generally speaking, when working with partial reloads and if you aren't getting the, result that you're expecting.
Trying to Take parts of the data model and try to find exactly where it's failing you. Kind of identify the root cost of the problem.
And then when you have that, we can go on to Qlik community and work on it together there. To kind of get it up and running.
Because it's So hard to. That question without knowing more details and also trying to find the exact point where it doesn't work.
Yeah. My basic troubleshooting brain says, you know, cut it in half, see if it works or fails and cut it in a half again, kind of happen till you've isolated the problem.
Yeah.
But. That's a little tricky.
I think, yeah, yeah, exactly. But they're, they're, I saw a follow up there from, the guy or, girl, the person who has the question, said that the problem is that the ride field sort of dropped on a partial reload.
There you go.
Yeah if they are what you could try to do is look at some kind of ETL structure where you see if you can load the app in multiple, multiple steps.
So, so you've given because there is a reason why you're doing a partial reload.
It's probably because it takes too long to do a full reload. You might just want to go in and check different specific parts of it and so on.
Hmm.
So in that case, trying to maybe create the application that just loads the derived fields and then generates, generates it to, to, the fields that you want to and then kind of connects that to the rest of the model, working with QVD files inside of.
Qlik Cloud or Qlik Sense wherever you're working on to kind of get that up and running.
That would be one, suggestion.
That's a good outside the box thinking. Thanks. Next question: is it possible to output the current filtering when exporting PDF?
Do I guess applying current selections?
So It's I think it's having that part of the Excel spreadsheet or something, to have you doing in QlikView.
Hmm.
Where at the bottom, we can says selections. I don't know if recurrence is port that.
Okay.
I think we are looking into it, but like that of my area. But just so that we export the report, you know what the year was selected, for example, that's including the data.
Yeah.
See what selections are applied. Interesting.
Next question: What are some of the other cool enhancements in the new pivot table and what version of the An on-prem will that be included?
So do you peel table if I remember correctly, was released in the bundle in the February release as a first one.
Of 2024?
It has, like, limitation. It didn't have number formatting. It didn't have some other things in there.
You had to use the most, There was an update to it in the May release that I'm not sure exactly what went in and in there, but I can show you what's live on cloud and if it's not in the May release, then it's in the November least.
Something to talk about also that we're moving from releasing 4 times per year to 2 times per year.
So I'm May and a November release. If I share my screen, you can look at some of the other things in here.
So the new pivot table; I would say there's a lot of inspiration from QlikView when designing it so you can put probably recognize the way that we're structuring and putting the measures in here.
So compared to the current one, we won't have that big labels up there with the live.
Pivoting and not showing the dimensional, in the correct place. Now we will actually have the dimensions being shown together where they are being used.
You can also interact with them. You can do searching them. You can do selections and things like that.
A big change is also that you can, make the width, not only work when it comes to the rows, but also for the columns.
So in this case here, I can make all these columns smaller. I think the minimum size is about 25 or 30 pixels.
So. I can make them uniformly equally large or small. If you don't wanna do this, interactively inside it in here, you can also Open up the settings for it.
And you can see that there are some Width settings for this, if I go to year and here.
Let's see. I can remember where it is. You can see here I have the column within here that I can specify as either pixels, fit the content, percentage of the screen.
Auto; trying to equally size it so that you fill up the entire space in here. So we can also use expressions to save it should be a small or a wide one.
So a lot more functions when it comes to the width of each column. So that's something we like, right?
What we also have in the new one is that you can say that you want the text to be represented as, for example, or the data.
In this case, the dimension could be text. It could be a link, so a URL to something.
You can also have images embedded inside the pivot table. Either S, measure or as, a, dimension, the, the row or the column level.
And these are things that are on cloud that I don't know if they made it into May release or not.
Otherwise they'll be in the next release afterwards. Thanks we are looking at is as I said before the heavy thing having an exploration panel for a period table where you can include and exclude dimensions and measures and feed them around.
We're all looking at having mini-charts in Pivot table. I can't see if we're actually going to do that or not.
This is something that we are looking into. And then of course a lot more when it comes to styling of the pivot table being able to say what is there.
Column background, what is the header background, things like that. Also one thing, I think it's for a table might be for pivotal also being able to use expression for if you want to have bold, it's italic on the line.
Similar to QlikView where you have font format expression. That we are working on.
When and so things get delivered I can't really say but yeah.
Thanks. We’re running pretty close to the top of the hour. One last question that came in was, can you talk about the best practices for the ETL process.
And Qlik Cloud and how to manage. Catalogs on an individual level, but I will see if we can
Yeah, I think that the last part of that share a source where you can acquire such knowledge is probably the best answer in this case because that's a big topic as well.
Yeah.
Yeah. But, so there, as you know, we're Qlik Talend, which means that we both have the Qlik side on the town side on the town side there as a lot of tools in place for working with.
With the detail process and getting data from, on from into the cloud. Much of their enablement material is on the a platform which we're calling Qlik learning right now which is gonna replace the old Qlik learning and be an aggregate of the information that was on that platform and also the information that was on Talend Academy.
So that will be the One place to go to get enablement on platform specific topics like these.
And that's a process that we're working on right now on consolidating those. As of right now, it's Qlik learning and it's Talend academy, but they're merging to the same platform.
And that's kind of where you should go to get that kind of, in depth, enablement and learning content.
Thanks. And I just wanna thank our wonderful panel, all these wonderful questions from everybody who's participated and I hope you all have a great rest of your day.
Bye bye.