Data Science

Unleash Python within Microsoft Power BI Desktop

We have all seen the amazing things Microsoft is doing each month with Microsoft Power BI updates. I’ve been working in the Microsoft ecosystem for 20 years now, and I’ve never seen an update frequency like the one we see with Microsoft Power BI and the Power Platform in general. These updates are not trivial either and they usually span across several areas of Power BI itself. With the recent August update, Microsoft has now giving us the ability to host, render and use Python script for visuals and within PowerQuery (M) for loading data frames into Power BI data sets. With this, we now have a great choice of either R Script visuals or Python Script visuals to help us craft more advance data stories. This includes such stories that would be descriptive in nature, but also moving more and more into the forecasting and predictive / prescriptive models. With this new Python visualization option, it is super important to understand now what does this bring to the table for crafting stories? What are the current limits? What is the perceived and known road map for Python being valid target for effective data stories in Power BI? To get this started let me state my sheer excitement around this news and release. I have used Python a good bit in my past endeavors. It is one of the most flexible languages in the wild, with the abilities to help craft data stories, mobile, web and general app development needs. It can execute and be hosted across all the major platforms, from Windows, Linux and MacOS. I would dare say, only JavaScript, holds a real candle to having such a wide range of useful targets to work in. Now lets dive into Python for Power BI, how to set this up and getting started within using it with Microsoft Power BI. In the following sections we will answer the questions set in the previous paragraphs which should enable you to start using Python and looking to incorporate it into your data story needs. (more…)

By brandon, ago
Datawarehouse

The arrival of the modern data warehouse – part i

One of the cornerstone technologies for enabling enterprise grade insights is the data warehouse. This has had many names adopted through the years, since the inception of the phrase itself by no other than Bill Inmon. (resource: Find out more on our resource page) Having coined said phrase back in the 1970's, with the goal in crafting a standard in which a companies data could be stored, accessed and used in order to provide a single source of truth in which describes and tell the story of said company. Thus the great data warehouse religious wars began that pitted two strong willed and minded camps. Those whom would embrace the original author of the concept itself found themselves flying and fighting for the Inmon camp. With this there was yet one more knight in which to battle until victory was claimed. If therefore you where not part of the Inmon camp then you found yourself digging in and amazed at the fact in which anyone would design for a company the construct that a data warehouse represents - no others than that of Ralph Kimball's approach to the design of the coveted single semantic layer. This, as time progressed onward, would then in turn be called an enterprise data warehouse. Inferring with its name the single and solely accepted version of a companies truth, and that Kimball on a tone of 7/8-1 would be used in designing a companies prized semantic layer.

(pictured above is an example of the Ralph Kimball data warehouse architecure)

(more…)

By Amy George, ago