Let's say I have a table with 100 different Jobs with all the Salaries and Years of experience. I'd like to do a linear regression on each of the different job. How do I use LINESTX in a smart way, avoiding to write 100 DAX queries? How can I link a dropdown list of the jobs, located in one tab of our dashboard, to LINESTX function? Is that possible? Basically how can I filter dynamically the table that is feeding LINESTX?
You could create a measure that calculates the expected salary for all the data. Use that in a report/page that has a slicer that is single select for the job titles. The user can chose a job title and then analyze the expected salaries based on the 'what if' slicer. Also, to create a line in a graph, you can create a variable inside the measure that gets the 'position' (e.g. years of experience, the X axis) and then use that in the result Y = VariableX * slope1 + intercept
I used COUNT function instead of SUM function as used since i am working with months to predict future sales as i have sales for 6 months. The parameter function did not work since it is factoring SUM instead of COUNT in the calculation
Well done fernan. great video to forecast easily
excellent explanation, thank you
I have been saddled with the task in my office to generate some predictive analysis in my report. This nailed it. Thank you so much Fernan
Hi Fernan, Is there any solution to add Polynomial Trend chart in a scatter plot. I searched over internet but didn't find any solution.
Microsoft excel office support has the best explanation of LINEST.
Thank you for the video, but i have a problem with my data when I add a column to the value section it will not work again. Please i need your help
Let's say I have a table with 100 different Jobs with all the Salaries and Years of experience. I'd like to do a linear regression on each of the different job.
How do I use LINESTX in a smart way, avoiding to write 100 DAX queries?
How can I link a dropdown list of the jobs, located in one tab of our dashboard, to LINESTX function? Is that possible? Basically how can I filter dynamically the table that is feeding LINESTX?
You could create a measure that calculates the expected salary for all the data. Use that in a report/page that has a slicer that is single select for the job titles. The user can chose a job title and then analyze the expected salaries based on the 'what if' slicer. Also, to create a line in a graph, you can create a variable inside the measure that gets the 'position' (e.g. years of experience, the X axis) and then use that in the result Y = VariableX * slope1 + intercept
I have Revenue and Cost in two different tables, how can I use the above to arrive at the Linear regression and predict for future years
I used COUNT function instead of SUM function as used since i am working with months to predict future sales as i have sales for 6 months. The parameter function did not work since it is factoring SUM instead of COUNT in the calculation