A Fine Slice Of SQL Server
Alex Velez lays out the case for tick marks:
Lately, I’ve noticed that more and more graphs don’t include gridlines. If it’s unclear, I believe this to be a positive trend. I, myself, rarely use gridlines, and often remove them when I find them in a graph I’m reviewing. But I don’t stop there.
More often than not, if a chart has gridlines, it will be lacking tick marks along the axis, and possibly an axis line as well.
Read on to see why.
Comments closedMarco Russo and Alberto Ferrari combine two common business requests:
In a previous article, Rolling 12 Months Average in DAX we showed you how to compute a rolling average over a time period. In this new article, we want to take you one step further and show how to compute a moving average over a certain timeframe, that takes into account only the working days. We present two variations of the same solution: one that is optimized, relying on a calculated column, and one that – despite being somewhat slower – works without requiring a calculated column. The latter can be useful in case you need to define the formula in a live-connected report, where calculated columns are not an option.
Because the formula needs to account for working versus non-working days, it cannot rely on standard time intelligence functions. Indeed, DAX time intelligence functions have no knowledge about what it means for a day to be either a working day or a rest day. The NETWORKDAYS DAX function would not be very useful in this case, because it would introduce a slow filter to compute the range of dates that includes the number of working days desired.
Read on to see how they solve this one.
Comments closedRoland Szirmai has fun with field parameters in Power BI:
Meaning that report users can switch between “dimensions” of the data. This is great and already provides a much better UI and UX, but there was no information about the limitations of what “fields” can you add to the parameter table.
To be more specific, I couldn’t find any limitation about adding measures (Explicit Measures) to the Field Parameter.
I think you can see where my mind wandered after that…
Read on for the result of Roland’s wanderings.
Comments closedDerek Jones digs into an image:
A month or so ago, I found a graph showing a percentage of PCs having a given range of memory installed, between March 2000 and April 2020, on a TechTalk page of PC Matic; it had the form of a stacked density plot. This kind of installed memory data is rare, how could I get the underlying values (a previous post covers extracting data from a heatmap)?
Read on for an interesting attempt at reverse-engineering the original numbers used to create an image. H/T R-Bloggers.
Comments closedRick Pack updates a package to support a project:
I am happy to announce a contribution to the biscale package that makes printing shorter labels using SI prefixes (e.g., 1,000,003 => 1M and 1,324 => 1.3k) far easier. This makes printing the legend in an attractive easier, although you can tell by the picture above that I still struggle with optimal uses of the cowplot package’s draw_plot(). I would love for the legend and map to be centered under the title.
The new si_levels argument for bi_class_breaks() takes a logical value of TRUE or FALSE for either a single or two-unit vector, with a single unit vector causing the specified value to be applied to both the X and Y variables. This matches Prener’s convenient functionality for the number of digits function dig_lab, as he requested in the Github Issue I created for this addition. Note that si_levels rounds the input number, if appropriate, based on the digits indicated by dig_lab, which defaults to 3.
Click through to get access to the update, as well as to see some of the visuals Rick put together with it.
1 CommentMeagan Longoria has a checklist:
For more formal enterprise Power BI development, many people have a checklist to ensure data acquisition and data modeling quality and performance. Fewer people have a checklist for their data visualization. I’d like to offer some ideas for quality checks on the visual design of your Power BI report. I’ll update this list as I get feedback or new ideas.
Read on for the list, as it’s a good one. For the most part, these also apply to visuals created in other tools.
Comments closedMichal Bar has a few updates to the Azure Data Explorer web tool:
We are focused on continuously improving the results exploration experience in ADX web UI, to make it easy and intuitive. Our goal is to provide an easy-to-use UI so that you will not be required to re-write KQL queries in order to perform light-weight data exploration.
Click through to see how you can search and filter within the results pane (something I’d like to see in other Microsoft data platform tools like SSMS), create series panels on charts from KQL, and more.
Comments closedAlex Velez takes us through stacked bar charts:
A few years ago, we posted a question on this blog that is as relevant today as it was years ago: “Is there a good use case for a stacked bar chart?”
Stacked bars are everywhere; you’ve likely seen them in a recent report, a dashboard, or in the media. Despite their prevalence, they are commonly both misused and misunderstood. In this guide, we’ll aim to rectify these mishaps by sharing examples, clarifying when you should (and shouldn’t) use a stacked bar chart, and discussing best practices for stacking bars.
Read on for plenty of good advice around when to use stacked (either regular stacked bar charts or 100% stacked), horizontal vs vertical, and how to format them when it does make sense to drop one in.
Comments closedGauri Mahajan looks at the key influencers visual in Power BI:
Once the Key Influencers are added to the Power BI report, it would look as shown below. The visual would be empty by default. The key areas that are required to make this visual works are Explain section and Analyze By section. The Analyze section is used to point to the variables or attributes that we intend to analyze. The Explain By section is used to point to the variables or attributes that may be influencing the attributes specified in the Analyze section.
I’ve found this visual to be pretty interesting if you have a good dataset.
Comments closedGauri Mahajan shapes the narrative:
To make it easier for the end-user, this job may be done by report or business analysts who may pre-analyze the reports, manually form textual narratives that summarize the key highlights in the report. While it solves the challenge in question, it opens a possibility of analysts’ bias getting introduced in the report, and the end-user may or may not agree with the narrative. Some systems solve this issue by employing complex machine learning / natural language processing / other artificial intelligence-based mechanisms to auto-generate smart textual narratives that summarizes the key highlights of the data. Though this approach works, it requires a significant number of resources and hard-to-find skills which is outside the bounds of a normal end-user who may want to use a reporting tool in a self-service manner and build a dashboard.
Modern reporting solutions like Tableau, AWS QuickSight, Microsoft Power BI, and others in similar league have been offering a feature to generate key insights using built-in AI/ML in the reporting tool which enables an end-user to extract insights as well as enables a report developer to have a smart visual that auto-updates the insights based on the change in the data.
In practice, this ends up being more of a fun toy than a really practical solution. Part of the issue is that decent analysis is hard, even more so when you have to develop something before even seeing the data or having any priors around feature importance.
Comments closed