Press "Enter" to skip to content

Month: November 2022

Power BI Model Documenter Version 2.1.0

Published 2022-11-08 by Kevin Feasel

Marc Lelijveld has a new version of the Power BI Model Documenter:

It is long overdue, but time for a new updated version for the Power BI Model Documenter external tool! The latest release lasts from the first week of 2022 and since then close to 6000 people have downloaded the Model Documenter installer directly from the website, excluding the downloads and clones from the GitHub repository. I can’t share enough how happy I am with all the interactions I’ve had with people from all over the world providing feedback, asking questions and willing to contribute to the Model Documenter.

Read on to see what’s changed since the last release.

Comments closed

DirectQuery Partitions on Hybrid Tables

Published 2022-11-08 by Kevin Feasel

Shabnam Watson noticed something interesting:

While I was tracing some Power BI datasets with Hybrid tables, I noticed that that the DirectQuery partition of a Hybrid table is used in all queries submitted to the Hybrid table, whether or not they are asking for data from that Direct Query partition. On first look, this does not look good because Power BI should be able to use partition pruning for a partitioned table, meaning it should only scan certain partitions. For example, if you ask for historical data from 5 years ago, the current day’s data is not needed to answer that query and hence, the Direct Query partition for today should not be used in the query.

But do read on for more information about what’s going on, as it seems there are layers of nuance here.

Comments closed

Accessing Google Trends Data from R

Published 2022-11-07 by Kevin Feasel

Sebastian Sauer looks at Google search data:

You cannot download as much data as you like, there are some restrictions, again, from the same source as above:

Google has incorporated quota limits for Trends searches. This limits the number of search attempts available per user/IP/device. Details of quota limits have not yet been provided, but it may depend on geographical location or browser privacy settings. It has been reported in some cases that this quota is reached very quickly if one is not logged into a Google account before trying to access the Trends service.[52]

Click through to see how you can access this data. In this case, the example focuses on specific categories but there’s a lot more within Google Trends.

Comments closed

Working with Multi-Channel Bots in Azure

Published 2022-11-07 by Kevin Feasel

Matt Eland creates a mega-bot:

The Azure Bot Service is effectively a registration for a conversational AI application on Azure. This registration allows you to connect a deployed chatbots to a wide number of supported channels that users can use to interact with the bot.

This lets you build one bot that can serve a variety of users across multiple different channels, including both text and voice channels.

Additionally, the Azure Bot Service gives you a centralized place to manage, secure, and monitor your bot, regardless of which channel people use to interact with your app.

Read on for an important caveat, as well as more information on Azure Bot Service.

Comments closed

Profiler Errors against Power BI Dataset

Published 2022-11-07 by Kevin Feasel

Shabnam Watson troubleshoots an issue:

I was trying to trace a dataset I had published to Power BI service using SQL Server Profiler and I was getting this error:

Either the trace with the ID of ‘MicrosoftProfilerTrace1667261566’ does not exist in the server with the ID of ‘autopremiumhostnorthcentralus001-081’, or the user does not have permissions to access the object.

Read on for the solution.

Comments closed

Moving Away from pg_dump

Published 2022-11-07 by Kevin Feasel

Pat Wright explains why pg_dump isn’t the best solution for backup and restore:

I’m still fairly new to Postgres having only started about 5 years ago. I started with Pg9.6 and we quickly moved that environment to pg10.   When I arrived at this company,  pg_dump was the only backup we were using.  The DB at that time was still around 50GB, it was reasonable to do the backups in a timeframe that worked for us.  A dump was done every night and stored off to another server.   After some time we started to test Pg_basebackup. This allowed us to full server backup each night. It was a huge improvement as far as speed and ability to handle much larger data sizes. 

Read on for the tradeoffs around tools and various thoughts from Pat.

Comments closed

Thoughts on Postgres Backups

Published 2022-11-07 by Kevin Feasel

Ryan Booz shares some thoughts on backups in Postgres:

To be honest, I feel pretty unqualified to talk much about backups in Postgres, partially because it’s been a couple of years since I’ve had to manage the uptime of an application. As PostgreSQL has grown in popularity and usage, there is a lot of effort being put into tools like pgBackrest, pgBarman, and some newer solutions like pgcopydb (ok, I realize this isn’t billed as a “backup” tool, but…).

What I can share are a couple of things I’ve learned about restoring clusters and databases, particularly given the extensible nature of Postgres and the variety of tooling.

Read on for thoughts on a variety of topics.

Comments closed

The Importance of Validating Postgres Backups

Published 2022-11-07 by Kevin Feasel

Grant Fritchey brings an important insight:

I’m very much just beginning my journey of learning PostgreSQL. I’ve been documenting that learning over here at Simple-Talk (more on the way there), including backups. For this post, I’m not going to tell you about my “experience” maintaining a PostgreSQL backup routine because, well, there isn’t any. Instead, I have something else to say about backups that I learned, the hard way I might add, while working in SQL Server, that is 100% applicable to PostgreSQL.

Click through for Grant’s thoughts. The “what” makes sense. The “how” is the important part.

Comments closed

Removing Backgrounds from Images

Published 2022-11-04 by Kevin Feasel

Brendan Tierney focuses on the subject at hand:

There are a number of methods available for preparing images for input to a variety of purposes. For example, for input to deep learning, other image processing models/applications/systems, etc. But sometimes you just need a quick tool to perform a certain task. An example of this is I regularly have to edit images to extract just a certain part of it, or to filter out all the background colors and/or objects etc. There are a a variety of tools available to help you with this kind of task. For me, I’m a Mac user, so I use the instant alpha feature available in some of the Mac products. But what if you are not a Mac user, what can you use.

I’ve recently come across a very useful Python library that takes all or most of the hard work out of doing such tasks, and has proved to be extremely useful for some demos and projects I’ve been working on. The Python library I’m using is remgb (Remove Background). It isn’t perfect, but it does a pretty good job and only in a small number of modified images, did I need to do some additional processing.

Click through to see how the tool works, as well as some cases it doesn’t quite get correct.

Comments closed