Post Reply 
Strange RREF (HP 50g)
10-04-2016, 09:44 PM
Post: #1
Strange RREF (HP 50g)
Hello HP 50g friends,

I watched a strange behavier of the RREF command.

If I input the matrix:

Code:
[[ 7.6440528636 7.6440528636 -300. -53.1753303963 ]
 [ -10.192070485 -10.192070485 0. 229.289735683 ]
 [ 61.847 0 0. 8. ]]

and use the RREF command, I get:

Code:
[[ 1. 0. 0. .129351464097 ]
 [ 1. 1.0192070485E14 0. -2.30608094922E15 ]
 [ 1. -.15384615385 4.61538461549E15 -1.82756879024E15 ]]

That is something else as aspected, but if I use a second time RREF I get the aspected form of the matrix:

Code:
[[ 1. 0. 0. 0. ]
 [ 0. 1. 0. -22.6262264632 ]
 [ 0. 0. 1. -.395973237876 ]]

Is there anybody who can explain me the strange behavier? It is not a question of the calculation mode (approxmative or exact)?

Thank you in advance,
sincerely peaceglue
Find all posts by this user
Quote this message in a reply
10-04-2016, 10:53 PM
Post: #2
RE: Strange RREF (HP 50g)
Hello, I don't really know what is happening, but you can do this: with the matrix in level 1, enter the command XQ, this transforms every decimal in the matrix to fractions and sets the calculator in exact mode, then the command RREF and then the command ->NUM.
Find all posts by this user
Quote this message in a reply
10-06-2016, 07:44 PM
Post: #3
RE: Strange RREF (HP 50g)
Hello Juan14,

it seems to a matter of exact or not, in AUR it is described to use RREF only in exact mode. It is a pity, because the exact mode is very slow.
Find all posts by this user
Quote this message in a reply
10-08-2016, 12:23 PM
Post: #4
RE: Strange RREF (HP 50g)
Hello Peaceglue, I was looking at your matrix, and it seems that all the numbers in the matrix must be all reals or all integers, and your first 0 in the the third row is an integer, this somehow force the calculator to go in approximate mode, change it to 0. and run RREF in exact mode and it works.
Find all posts by this user
Quote this message in a reply
10-09-2016, 09:31 AM
Post: #5
RE: Strange RREF (HP 50g)
Hello Juan14,

I replaced the approach using RREF with the Cramer's Rule and get for over 600 solvings of systems with three unknowns (inclusive checking wether DET is zero or not) a calculation time of ca. 6 minutes (with RREF I get 4 times more). That is sufficient fast enough for my purposes.

But anyway, thank you very much for your help.

Greetings
peacecalc
Find all posts by this user
Quote this message in a reply
10-09-2016, 01:01 PM
Post: #6
RE: Strange RREF (HP 50g)
Hey folks,

the hp 50g is a real number crunching machine. Without the DET function but using CROSS and DOT the duration is only less then 4 minutes. wow!

Greetings peacecalc
Find all posts by this user
Quote this message in a reply
Post Reply 




User(s) browsing this thread: 2 Guest(s)