Ask Your Question
0

mdeterm solving matrix determinant [closed]

asked 2018-05-04 13:57:09 +0200

OSS-user gravatar image

updated 2018-05-04 21:54:40 +0200

Hi, I've stumbled accros with a strange issue, using LO Calc in spreadsheet calculating matrices, especially a determinant of the matrix with mdeterm() function. The function gives wrong answer. So here is the data in the calc file: https://drive.google.com/open?id=1S4p...

Octave gives answer for this matrix Det(A)=23503.602875 (matrix for solving in octave: A=[10, 22.5, 71.25; 22.5, 71.25, 253.13; 71.25, 235.13, 958.31]) error in third row a32 index

I've checked with wolframalpha and gives the same answer as Octave.

I found similar issue with this function in the buggzila . But it is shown, that this bug is fixed.

So, why do LO Calc gives me wrong answer? Any suggestions?

P.S. LO version: 5.4.6.2 (x64)

edit retag flag offensive reopen merge delete

Closed for the following reason the question is answered, right answer was accepted by OSS-user
close date 2018-05-04 22:01:37.228160

2 Answers

Sort by » oldest newest most voted
1

answered 2018-05-04 17:09:35 +0200

Regina gravatar image

I find no error in the calculation in Calc. Do you have considered, that the displayed values are rounded? If you compare it with other calculators you have to use

  10       22.5      71.25
  22.5     71.25    253.125
  71.25   253.125   958.3125    

det(A)=6806.25

If you do not believe it, calculate it with Sarrus-rule with a pocket calculator

edit flag offensive delete link more

Comments

With the example document given and Sarrus rule in Calc:

=A4*B5*C6+B4*C5*A6+A5*B6*C4-C4*B5*A6-B4*A5*C6-C5*B6*A4

=> 6806.25

It's "funny" but sad that both Octave and Wolfram apparently give the same wrong answer.. yet other calculators on the net agree with Calc.

erAck gravatar imageerAck ( 2018-05-04 20:13:58 +0200 )edit
0

answered 2018-05-04 21:48:55 +0200

OSS-user gravatar image

updated 2018-05-04 21:50:39 +0200

Hi, Regina,

thanks for the answer. That definitely has an influence on result, but not to much. The difference is by few tenth's. So, yeah, roundof is, because I didn't check what real value was in a cell e.g. 958.3125 Calc shows in general format as 958.31.

But real issue was human failure! When typing matrix, I've accidentally messed up one number in third row i.e. it should be 253.13, and my typo was 235.15. So Octave gave me another result.

Again, everything works great, all programs shows the same answer! Sorry for false interpretation.

P.S. I've checked the determinant with Sarrus rule, and the result was again - correct. Thank you!

edit flag offensive delete link more

Comments

@OSS-user - You have "closed" this Q&A, but you haven't "accepted" the "right" answer -- you do that by clicking the symbol at the top-left of the correct answer. Please do this! That's a signal in the Askbot systm to other users that the question has been solved. Thanks.

David gravatar imageDavid ( 2018-05-04 23:02:50 +0200 )edit

Question Tools

1 follower

Stats

Asked: 2018-05-04 13:57:09 +0200

Seen: 182 times

Last updated: May 04 '18