When you should avoid using bidirectional filters in Power BI

Поделиться
HTML-код
  • Опубликовано: 26 июл 2022
  • Bidirectional filters are a powerful tool, probably too powerful for most data models. In this quick talk we show the reason why most users should avoid using the feature, unless they understand well the implications.
    Speaker: Alberto Ferrari SQLbits.com/speakers/Alberto_...
    Speaker Blog: www.sqlbi.com
    Speaker BIO: Alberto started working with SQL Server in 2000 and immediately his interest focused on Business Intelligence. He and Marco Russo created sqlbi.com, where they publish extensive content about Business Intelligence.
    SQLbits.com/Sessions/When_you_...
    Tags: Azure,Testing,Power BI,Modern Analytics,Modern reporting and insights,On Premise,Intelligence and Analytics,Developing,Successful Delivery,On Premises
  • НаукаНаука

Комментарии • 4

  • @anissahli5698
    @anissahli5698 2 месяца назад

    Juste excellent !!! Tout simplement excellent merci à vous !

  • @laurenceabrahams3996
    @laurenceabrahams3996 Год назад +1

    Finally the answer that makes sense, thank you. Fellow trainer

  • @OneNI83
    @OneNI83 9 месяцев назад

    Nice.Can we know the DAX that has been used and the which methods leads to the correct answer on that example?

  • @Wzxxx
    @Wzxxx 5 месяцев назад

    What doeas it mean that “the path is preffered” . It just doesn’t go both sides simultaneously ? It cannot or what?- it pretty logical and intended sometimes. I completely don’t understand the idea behind ambiguity and the whole “danger”. No one really explains what is this. What is wrong when filters comes from both sides from date? Or maybe it just can’t? No one explains it? At the end it seems that it filters different attributes in fact table, or not? Really badly explained. I am stuck at real problem with my model but this explanation seems really bad…like why the filter cannot just come from both sides?. If u filter date it filters sales in selected date and product purchased in selected date which then filters filtered product sold or not? So at the end it simultaneously filters different attributes - dates and products (filtered by same date of purchase). Why it seems to be wrong? This is logical. If I add a different date table and select same date and filter purchases will it make any change at the end? The idea behind the final filtering is the same.