Joleen Bothma

I'm now offering specialized Power BI consulting services! 💥 🎉 If you're a small business owner interested in using data to achieve your business objectives, the best place to start is business intelligence. DM me to chat about how I can help you do that. #PowerBI #Analytics

I'm bringing back read-only Fridays. I'll be spending the day reading, learning, and exploring new things. Complete guilt-free permission to go down the rabbit hole. Today: the Poisson and Negative Binomial probability distributions for analyzing count data.

MIT researchers have created GenSQL to analyze data in a database. The goal is for users to query their data without having to know all the details. What could go wrong, right? It sounds cool but I'm not sure if half-baked, uninformed analyses are the way to go...

Do you have your own data cleaning or data exploration pipelines? 🧐 I'm thinking of building a pipeline containing the most frequent (and obviously most repetitive) tasks I do every time I explore a new dataset.

TIL: when looking for missing data in python, you can pair .isna() with .any() to quickly see if there are any missing values in each column in a pandas dataframe.

AI and LLM's have interesting and immediate applications in finance and accounting. For example, AI can automatically: 🤖 categorize transactions 🤖 detect errors or fraud 🤖 perform financial analysis But be careful! Understand where your data is going first!

Don't get stuck using a particular tool. Always be ready to adapt and change. I learned Qlikview in my first job, and within 3 months, I had to start learning Power BI. FAST! The entire tech ecosystem is in a constant state of flux. Move with it.

Data visualization tip: Don't let colors become noise. Choose colors wisely and with purpose. Add a splash of color to draw interest or to highlight critical data.

Recently discovered TimeGPT - the first pre-trained foundation model for time series analysis. It can produce accurate forecasts for new time series without the need for training. It's still in beta and isn't open source, but it marks a major milestone in time series analysis.

If you feel overwhelmed when starting a new data science project, do this: Keep asking yourself, "What's the next step?" until you get to a step that feels easy. Scott Young recommends this strategy to improve focus. The idea is to break down a task until it's easy to do.

@adild2k Sure, you're right. But I wanted to address why simply setting a many to many relationship is generally considered a bad idea unless you know what you're doing and what the consequences could be

@joleenbothma Thanks for sharing, but how about the resolution? Any intermediate table between customers and products with a 1 to M relationship would able to address these questions

Why are many-to-many relationships in Power BI a total nightmare? If you've ever been baffled by the constant warning to avoid them, I'm here to shed some light 👇 #powerbi #datascience #datavisualization #dataviz

I hope you've found this thread helpful. Follow me @joleenbothma for more. Like/Repost the quote below:

Ultimately, unless you are very familiar with how the many-to-many relationship affects your dashboard, it's advisable to steer clear of them. Personally, I think they're more trouble than they're worth and the last thing I want is for my dashboards to lose credibility over it.

1. It can get complicated Answering simple questions like "which products did customer A buy?" is complicated since there are multiple entries for customer A and for the products. So we need to be specific about which transaction(s) we are talking about. It's not obvious.

Consider this example: You have 2 tables in your data model: customers and products Each customer can purchase multiple products, and each product can be bought by multiple customers. This creates a many-to-many relationship. What's wrong with that?

I've noticed a pattern with AI-written content: Title Case. Title Case Everywhere. Us humans are far too lazy to use that much title case in 'normal' writing. Have you noticed any patterns that make you suspect something was written by AI? #chatgpt #ai

You can build a full-stack application using Python alone. You don't need JavaScript, CSS, or HTML. If you are a data scientist or someone dealing with data, here is an open-source Python library that will let you build end-to-end production applications without worrying about learning web development: github.com/Avaiga/taipy Star the repo! Taipy is a Python library. It has a library of pre-built components to interact with data pipelines, including visualization and management tools. It supports tools for versioning and pipeline orchestration. And it's open-source. And it comes with a Visual Studio Code extension that will get you started without writing any code. Thanks to the team behind Taipy for collaborating with me on this post. Adding this to your tool belt is one of the easiest ways to 10x your Data Science career in 2024.