Just don't ignore the economics of productivity. We need a world where we can produce immense amounts of food and energy with as little impact to the environment as possible.

As a data professional for 25 years AI is such a valuable tool. It increases my productivity by 10x, soon to be 100x. Increasing productivity is how humans can produce food cheaper, create new innovations (energy efficient transport, solar), and will ultimately be a great benefit to society.

No doubt that AI isn't going away even if the AI bubble pops. We need to learn to to measure carbon and create an economy that can exist with reliance on growth. (And distributes income more widely)

Yes, ingest each file individually. Don't navigate the SharePoint file list.

A few reasons that are common are 1) Slower performance 2) Dim table gets limited by a fact when it shouldn't 3) potential for ambiguous relationships when expanding (can't create new relationship that is needed because another path exists)

It's certainly less risky in a small model. The larger a model gets the more important data modeling becomes! Starting with a few rules makes life easy. 1) Dims filter Facts with a one way relationship. 2) Facts don't filter other Facts. 3) When Facts need to filter Facts use a SCD or bridge table.

Got it. Expose those fields via measures when possible. If the employee fact must filter other facts then decide if that table is actually a dim. The fact could be duplocated to move the attributes you need into a dedicated dim table with new primary key. Similar to a SCD with surrogate key. 😅

Other measures should work perfectly if the attributes are all from your common dimension tables. Adding attributes from other fact tables is where it becomes challenging.

Adding an attribute from your current Employee table (which is acting like a fact table) will cause duplication. This can be avoided by only exposing data from Employee Fact via a measure.

Thanks for sharing data model. The measure is the glue. Because blank measures results are filtered automatically In Power BI. Create a visual with EmployeeName from EmployeeDim, and Leave Days on rows. Then click a single date and the visual will show only Employees with Leave Days on that date.

The typical best practice is to duplicate the Employee table to create "Employment Fact" (limit columns to just the date fields needed), then use that data for employee measures. This insulates Employee and Date to keep them as fully independent dimensions.

It's 100% possible to build a great data model and never enable a relationship where a Fact filters a Dim. There are some cases where facts can filter other facts but this needs to be done very carefully. The measures are the glue that bind everything together to create the effect you are after.

Enabling this bidirectional relationship will impact the future expandability of your model. When you have a new fact table to add, like "Employee trainings completed", you'll need a relationship between Date and Employee to this new fact table. However the two way relationship creates two pathways

Typically the measure "Leave Taken" would be used to show the intersection of Date and Employee without create the two way relationship. Add this measure to any visual and it will narrow the data. The measure can be in a hidden column or even visual filter. I'll explain why in my next reply.

Please add "Goth art, Anime, Dragons, etc."

Now turn those circles into three circular tilted pyramids that converge at the top. Utopia may not exist but we can get closer through innovation.

First we have to figure out the energy and climate problems we have, but once that is solved then why not? In the future I'd love to have a personal assistant to download my music, find me new music, keep tabs on the bands on town, prep birthday cards for personalization... things I don't do today.

Slow and consistent change in progress.

You can ask chat GPT to write a .bat command or powershell script to move the files after they are unzipped.

Small stepping stone in the journey to a code based approach. Soon these users will realize Excel is holding them back and the power of AI code generation will win.

Great analysis and take. I'm loving the ability to create technical products more quickly than I could have ever done before! While AI has caused a lot of disruption for me (budget shifts / job stability), it's opening up a lot of possibilities.

Use the chrome translate feature to check out this great data visualization. Increasingly AI will help analysts produce data stories like this.

It does encourage some people to work on their resume & get in a habit of applying, but if there aren't enough jobs then it's all pointless. The next wave of robotic and AI automation is here and we need policies that help people.

Jobs are on the decline. Especially simple jobs. Examples: Automated driving, automated AI code fixes, self check out lanes, call centers using AI.

GitHub Copilot: The agent awakens Introducing agent mode for GitHub Copilot in VS Code, announcing the general availability of Copilot Edits, and providing a first look at our SWE agent.

GitHub Copilot introduces agent mode! The new age of the digital creator is in full swing.

github.blog/news-insight...

Eventually the benefits of the onslaught of Copilot integrations will be productivity enhancing. Users need to be a part of that change by providing feedback and sharing features that work. Change is hard, and somewhat painful, but I have no doubt Copilot integrations will be awesome soon enough.

Have you tried VS Code edits to add new features to the Python script? It's so awesome.

Git Hub will always have more code to train GenAI.... Microsoft and Open AI have a lead in that department!

Julius AI | Your AI Data Analyst Julius is a powerful AI data analyst that helps you analyze and visualize your data. Chat with your data, create graphs, build forecasting models, and more.

AI tools will change the way data analysis is performed by 180 degrees. Soon asking a data analyst to access data, prepare queries, and format results will be a thing of the past. Natural language queries and copilots will perform all these steps and deliver data before you ask. julius.ai

Introducing Copilot Edits Copilot Edits allows you to get to the changes you need in your workspace, across multiple files, using a UI designed for fast iteration. You can specify a set of files to be edited, and then use natu...

SQL is on the way out. AI generated Python is the future. AI code can more easily follow the patterns of millions of lines of code. Who wants to code SQL in an IDE without Copilot? Nobdy. code.visualstudio.com/blogs/2024/1...