Qlik Python ShowCases

[Music] hi guys this is Christopher earthquake some people claim that my entire household is totally clique branded well that's a rumor let's say they have a look together in two clicks integration possibilities with Python let's first look at the front-end integration in order to this to work I've got Python installed on my machine I'll start a command prompt and I can set my working directory accordingly I don't want to claim this for myself this is really great work done by Riley McDonald my colleague and let's start the Extension Service this is using its service on on the given port and waits for incoming Qi RPC calls so ygr pc g RPC is the protocol used by the server side extension if you don't know what server-side extensions are here is the link to check this out I need to open your settings file in of clixsense desktop you can find it in this place and make sure you have this key SSR plug-in equals so at least what you needed to do here is put this into your ne file because I have other examples I have a semicolon still separated list with some other options as well and now we need to start or restart clixsense desktop and now something is printed here in the python console it added the capability linear regression so now it's a registered as an external function and as i log in now in clixsense desktop I can go to the example and I can see the linear regression line calculated by python it's the red line here if we quickly inspect what's happening there basically a function called Python regression to be routed through the given Python endpoint and calling the function linear regression it gives the some sales to the Python codes together with the dimensionality given here in other words for each year month combination it's sending a sum and it's receiving the linear regression for this and of course this is dynamic so if I change my selection and get a smaller part of it it's making a different linear regression accordingly this is real time you can see that each time I change the selection it prints that the function has been executed so the other example of how this can work is through the script let's stop this one go to a different project I mean you could start them in parallel if you want it and start the main code of it it's similar as before just a different port it's waiting for the suicide extension to register and since it's click since desktop it will not automatically do that so I have to quit and restart click since desktop and we can keep watching the window what's going to happen as soon as click sense comes up as soon as it comes up it registers the new function and this time I'll open another app and this is the food analytics app it's about sales in a retail shop with categories of products called departments here then we have products and this translates into sales and we know from a shopping basket ID which products were brought together in one basket now I would like to apply a advanced algorithm that calculates which products are bought together of course it would take much more time to explain how this has been done but essential parts are seen here in this script editor where I first load my data from flat files in this case Q V DS my dimensions my facts which are the orders and via that order ID I can determine which products were bought together in a basket so here this part is about sending data to Python and loading the results so it's sending free fields to Python the order ID the Product ID and purchased which has been prepared in memory before and it will receive five columns back which is again the product ID the other product IT support confidence and lift these are free scores that are private to the specific algorithm that we are calling and if I execute the script the first part is no miracle it's just loading a lot of data and now it's called the function and it's waiting for the result now it's a patents drop to do something so the script is holding here Python is heating up my processors on my notebook and in a little while we will see results streamed back so I've received back a couple of lines and this has become part of my data model let's quickly have look so this is where the result has been persistent it is a new table that's hanging between the product ID and a new table called other product ID and now you can do the detailed drills of that data set so that's demonstrate is here I do the product name I do the other product name and and the confidence score and as we do so we can immediately see the products that are most confidently bought together so it's a bag of organic bananas together with Hass avocado you can of course now see both scores greater than point free with just a single click and there you go that are in a nutshell two modes of interacting with Python the life mode and the script mode both serving different use cases so with that come to an end of this video today hope you like it and if you do so please subscribe to my channel as I always love seeing new members joining the club see you next time [Music]

Channel: Christof Schwarz

Views: 20,979

Keywords: qlik, qliksense, Qlik, Qlik Sense, python, Python, QlikSense, integration, Integration, serversideextensions, server-side extensions, SSE

Id: nI-Bgal-dSM

Length: 8min 7sec (487 seconds)

Published: Sun Jan 27 2019