HELP HP 48G "INV ERROR"
Kam Wai Lau
I have a HP48G and I was doing some simple matrix calculations. I did a
3x3 matrix inversion and I got the error:
INV ERROR:
Infinite Result
The 3x3 Matrix is:
[[4 5 6]
[7 8 9]
[10 11 12]]
The puzzling thing is my friend's 48SX can do the inversion whereas mine
cannot.... Anyone can help to solve the puzzle?
Thanks
Kam
Iceman
Your friend must have typed it in wrongly. My S gives the same error. The
DET of the matrix is zero and that's why the matrix doesn't have an inverse.
IF DET(A)=0 THEN A has no inverse.
___ /---\
__|_|__ "I may be a coward, (\ /)
(o o) but I'm a greedy little coward" -----ooO-"--Ooo-----
( - ) l.r.sluijk@student.
( : ) utwente.nl
( : ) --------------------
(_________)
"Be there or be somewhere else..."
Billy Allen
: Hi there,
: INV ERROR:
: Infinite Result
: The 3x3 Matrix is:
For starters, that matrix is singular, hence the error. As to your friend
the hp48 G series is much more accurate with matrix calculations than the
S series. It is thus possible that the S inverted a reasonable facsimile
of the matrix, or due to numerical nonsense that is irrelevant, produced
something that is close to an inverse. See GD 9 matrix.doc for more info
on S vs. G matrix performance.
Billy C. Allen, III
bca...@tree.egr.uh.edu
Steve VanDevender
Hi there,
I have a HP48G and I was doing some simple matrix calculations. I did a
3x3 matrix inversion and I got the error:
INV ERROR:
Infinite Result
Not all matrices are invertible. Matrices with a determinant of
0 are not invertible.
The 3x3 Matrix is:
[[4 5 6]
[7 8 9]
[10 11 12]]
Your matrix has a determinant of 0. Try using Gaussian
elimination to find the inverse; you won't be able to because the
rows of the given matrix aren't an orthonormal basis.
The puzzling thing is my friend's 48SX can do the inversion whereas mine
cannot.... Anyone can help to solve the puzzle?
My HP 48SX gives the same error as your 48G. Perhaps it is
related to numeric precision settings. I generally use STD
rather than FIX or SCI. Because of limited precision, some
matrices that should have a determinant of 0 actually get a
calculated determinant that is near 0 instead. Such matrices
will also be inverted without error, although the inverse matrix
is essentially useless.
The HP 15C Advanced Functions Handbook has an excellent
discussion of error analysis for matrix operations.
Carsten Witzel
: INV ERROR:
: Infinite Result
: The 3x3 Matrix is:
: Thanks
: Kam
Could you please think about the mathematics involved before posting questions
like that? The above matrix has det=0, so there does not exist an inverse!
Your friend probably has set the flag which causes the HP48 to say 9.999..E499
instead of "infinite result". So his "solution" is nothing but a rounding
error.
--
Carsten Witzel -- E-Mail: wit...@hp.rz.uni-duesseldorf.de
Mettez votre nom ici
Your matrix CANNOT be inverted,because its DET is equal to zero!
On your friend's calculator, you haven't typed the same matrix, certainly.
Mathematically speaking, the third row can be obtained with
2 x row2 - row1 = row3
That is the reason why.
Try any (almost!) other matrix and you will be proud of your 48G again!
Bye!
Mahtieu Besson