Calc: Linear Regression output not useful in 6.4.7.2

Hi.
I’m new to this forum though not to LibreOffice. I used the linear regression long before and it resulted in cells giving a slope and an offset. This allowed to draw a line and to extrapolate. However, this seems to be gone. The output generated as shown below is maybe statistical useful, but it does not allow to draw a line and to extrapolate. Is this intended or am I missing something?

I am running CentOS 8 as a desktop and the LibreOffice version shipped from the repositories is 6.4.7.2 .

Thanks for any hints.

Please update first and check if the error is fixed, thank you.

The very first step: reproduce bug using latest fresh/still version

Download LibreOffice


Also check:

Computation-related issues in Calc ( [OpenCL]

Thank you.

Thank you for looking into this. I did not expect a bug. I wanted to understand what happens. I clearly did not. And for the time being I prefer to stick with what is in the repositories, in this case for CentOS 8 Stream.
I agree, if this was for a bug, the test should be made against the latest version.

You could at least have told us what you did (and screenshots aren’t much helpful to recreate, a sample document is better). Apparently you applied Data → Statistics → Regression… on cell range A3:B6. Version 6.1 had a very limited output set that explicitly mentioned Slope and Intercept. Anyway, the values are in the LINEST results, in your example slope is in A12 and offset / intercept / intersection with the y-axis is in B12. You’ll find them also as Coefficients for Intercept and X1 in cells B34 and B35. See also LINEST help.

Thanks for picking this up! And yes, next time I’ll provide a document.

Yet you exactly figured what I did: Two columns of Data, Linear Regression applied exactly from the menus. And then the screenshot is the output. Everything by the books.

In this version of Calc, the help still suggests the old output with slope and offset. And at this time I was at a loss. I did not even figure that LINEST is a function. It was just one part in an enormous clutter of numbers I did not expect and did not see documented.

And so the second part of my last but one sentence applied: “…did I miss something?” Yes, and you correctly pointed me to the single two numbers in this out put that are most important to me: A12 and B12, which you could tell from the screenshot.

Thanks indeed for kindly looking into this an providing the solution!

PS: To whom it may concern: The call to this linear regression provides two columns titled “Regression Statistics”. The respective values are preceded with the names of the quantities they represent. I think, this is how it should be. I do not know, in which context LINEST ist usually used, but preceding the number fields in the output of LINEST with the names of what these numbers are, would be desirable I think.
PPS: I seem to be out of luck to get paragraphs. … still learning.