Search tips

Search tips

# Discussion Groups | Matlab DSP | what's the smallest real positive number in Matlab?

Technical discussion about Matlab and issues related to Digital Signal Processing.

Is this thread worth a thumbs up?

0

# what's the smallest real positive number in Matlab? - Lan Wang - Jul 4 1:12:00 2003

 Hi Folks, As a new user of Matlab, I'm puzzled by the following two questions: 1. I typed "realmin" in the command window and it gave me 2.2251e-308, but why did I get a matrix with a determinant equal to 5.434722e-322? I used "fprintf("%e", d)" to print the determinant. 2. In my result, DETERMINANT = 0.000000E+000 occurred for a matrix, which is zero in Matlab (and I verified it), why was there no error message, nor warning, coming up when the matrix with this determinant was inverted in my program? (But in the command window, when I tried to invert an obvioulsy singular matrix, the error message "Warning: Matrix is singular to working precision." showed up.) Many thanks. Richard

# Re: what's the smallest real positive number in Matlab? - Nandakishore Ramaswamy - Jul 6 0:48:00 2003

 Regarding your 2nd question the warning does show up but the answer comes up as "Inf" which I suppose is infinity. -Nandakishore Lan Wang <> wrote:Hi Folks, As a new user of Matlab, I'm puzzled by the following two questions: 1. I typed "realmin" in the command window and it gave me 2.2251e-308, but why did I get a matrix with a determinant equal to 5.434722e-322? I used "fprintf("%e", d)" to print the determinant. 2. In my result, DETERMINANT = 0.000000E+000 occurred for a matrix, which is zero in Matlab (and I verified it), why was there no error message, nor warning, coming up when the matrix with this determinant was inverted in my program? (But in the command window, when I tried to invert an obvioulsy singular matrix, the error message "Warning: Matrix is singular to working precision." showed up.) Many thanks. Richard Yahoo! Groups SponsorADVERTISEMENT _____________________________________ /groups.php3 Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

# Re: what's the smallest real positive number in Matlab? - Pavel Jiranek - Jul 6 15:00:00 2003

 Make sure if you're working with single or double precission. Paya Lan Wang wrote: >Hi Folks, > >As a new user of Matlab, I'm puzzled by the following two questions: > >1. I typed "realmin" in the command window and it gave me 2.2251e-308, > but why did I get a matrix with a determinant equal to > 5.434722e-322? I used "fprintf("%e", d)" to print the determinant. > >2. In my result, DETERMINANT = 0.000000E+000 occurred for a > matrix, which is zero in Matlab (and I verified it), why was there no > error message, nor warning, coming up when the matrix with this > determinant was inverted in my program? (But in the command > window, when I tried to invert an obvioulsy singular matrix, the error > message "Warning: Matrix is singular to working precision." showed > up.) > >Many thanks. > >Richard

# Re: what's the smallest real positive number in Matlab? - navaneetha krishnan - Jul 7 18:34:00 2003

 Regarding question 1, it is possible to have values less than realmin. These are called denormals and can be as low as eps*realmin. See http://www.mathworks.com/company/newsletter/pdf/Fall96Cleve.pdf for details. On question 2, make sure warnings are not switched off in your program. Navan --- Lan Wang <> wrote: > Hi Folks, > > As a new user of Matlab, I'm puzzled by the > following two questions: > > 1. I typed "realmin" in the command window and it > gave me 2.2251e-308, > but why did I get a matrix with a determinant > equal to > 5.434722e-322? I used "fprintf("%e", d)" to print > the determinant. > > 2. In my result, DETERMINANT = 0.000000E+000 > occurred for a > matrix, which is zero in Matlab (and I verified > it), why was there no > error message, nor warning, coming up when the > matrix with this > determinant was inverted in my program? (But in > the command > window, when I tried to invert an obvioulsy > singular matrix, the error > message "Warning: Matrix is singular to working > precision." showed > up.) > > Many thanks. > > Richard __________________________________