Mailman 3 March 2006 - SciPy-User

Help required regding Eigenvectors
by Sudheer Phani March 14, 2022

March 14, 2022

Hello All, I have been encountering some strange problems with the calculation of eigen vectors using python. The problem is with the sign of the eigen vectors calculated. On Very few occasions the eigen vectors calculated using python have an opposite sign than the actual eigenvectors, though the magnitude is correct. Can any one of you plz help me in this regard and suggest me some solution if you have encoutered similar problems . Any help in this regd is appreciated Thanks in Advace regds Sudheer

5 5

3D plotting
by Robert Cimrman Sept. 10, 2009

Sept. 10, 2009

Hi, I need to do some simple 3D plots, so I tried the two Cookbook recipes (xplt, mplot3d) and failed: * xplt: does not work when using a custom installation directory such as python setup.py install --root=/home/share/software apparently the gistdata directory path is wrong: gist.error: failed to create drawing -- Gist work.gs style sheet missing * mplot3d: does not work with my version of matplotlib (0.80). I have made the changes mentioned in the Cookbook to no avail. (Axes.__init__() args apparently changed, as well as some other matplotlib object attributes) Any ideas? mplot3d looks great, I would really like to use it! thanks, r.

6 6

Floating point exception in scipy
by Rob Hetland April 3, 2006

April 3, 2006

Compiled on an Intel Mac os X using gcc 4.0.1 (the only one available on intel macs) and gfortran (from hpc.sf.net). Python is MacPython Universal build 2.4.3. Numpy compiles without a hitch, and tests with no errors. SciPy also compiles without errors, but I get a floating point exception when trying to test scipy. This includes doing scipy.test (10,10). Below are the details of some attempts. I'm really not sure where to begin, as it compiles fine. Jordan Mantha has claimed success basically following the PPC build instructions with this compiler configuration, but I have not had any luck. I have also tried to exclude modules that need umfpack, but that also failed in a similar way. Any ideas? -Rob mire:~/src/python$ python Python 2.4.3 (#1, Mar 30 2006, 11:02:16) [GCC 4.0.1 (Apple Computer, Inc. build 5250)] on darwin Type "help", "copyright", "credits" or "license" for more information. >>> import scipy import linsolve.umfpack -> failed: No module named _umfpack >>> scipy.test(10,10) Floating point exception mire:~/src/python$ python Python 2.4.3 (#1, Mar 30 2006, 11:02:16) [GCC 4.0.1 (Apple Computer, Inc. build 5250)] on darwin Type "help", "copyright", "credits" or "license" for more information. >>> import scipy.special import linsolve.umfpack -> failed: No module named _umfpack Floating point exception mire:~/src/python$ python Python 2.4.3 (#1, Mar 30 2006, 11:02:16) [GCC 4.0.1 (Apple Computer, Inc. build 5250)] on darwin Type "help", "copyright", "credits" or "license" for more information. >>> from scipy import * import linsolve.umfpack -> failed: No module named _umfpack Floating point exception ----- Rob Hetland, Assistant Professor Dept of Oceanography, Texas A&M University p: 979-458-0096, f: 979-845-6331 e: hetland(a)tamu.edu, w: http://pong.tamu.edu

3 2

Filtering high frquency noise
by Yannick Dirou April 3, 2006

April 3, 2006

Hello, I have a multirate signal (from 18 to 24 sample per day), and if i plot it i see something like "high frequency" noise (actually not that high but higher than the remaining of the signal), i thought about using a median filter but this is not good for a multirate signal, then i though i could use scipy filters to do the job, unfortunately i know nearly nothing in filter design, and don't know how to do the job. Is there a tutorial or simple example to design a low pass filter? the signal data is made of a datetime in epoch format and the measured value. Thanks in advance, Yannick

4 3

Re: [SciPy-user] maxentropy
by Matthew Cooper April 2, 2006

April 2, 2006

Hi Ed, Thanks very much for your reply. I think it helped a lot, but I may be a bit confused about conditional versus unconditional modeling. What I'm doing is similar to text categorization. I observe some text (vector X) and want to determine if a binary label (scalar y) should be applied (y=1) or not (y=0). So, I look at this as using maxent to estimate P(y|X). In this case, is my sample space simply {0,1} or is it the space from which X is sampled? I had thought this was conditional modelling, but I don't want to explicitly train models for every different X, rather I want to select and weight features from the whole corpus that in turn imply, for each X', some P(y|X'). I assumed it was conditional because I'm not modelling P(y',X'). I have been trying to build models using a sample space with tuples as elements like (X_n,y_n). I then am greedily building a set of feature functions using information gain to select features. It's not quite working, and I'm worried I'm not defining the sample space properly. I will try to figure out how to define the sample space as {0,1} and then redefine the features, but I'd appreciate any advice. Also, if I get this working, I'd be happy to try and help with adding some feature selection examples to your documentation. At the moment, I'm still, as you can tell, figuring out what I'm doing. Thanks, Matt

2 9

Re: [SciPy-user] select
by Pierre GM March 31, 2006

March 31, 2006

Lionel, Select is equivalent to if condition1: choice1 elsif condition2: choice2 else: default In your second example, if condition1 is false , then condition2 is always true: ( ! t>90 => t<90 => t<300). Same thing if you reverse the order of the conditions, of course. I find the base functions greater, equal, less... quite useful: select([greater(t,90)*less(t,300)],[t,]) is a good start.

1 0

select
by Lionel Roubeyrie March 31, 2006

March 31, 2006

Hi all, I have a little question about the select function, how can I select items between values: ####################################### lionel[~]58>t Sortie[58]: array([ 0, 18, 36, 54, 72, 90, 108, 126, 144, 162, 180, 198, 216, 234, 252, 270, 288, 306, 324, 342]) lionel[~]59>select( [t<90, t>300], [t,t] ) Sortie[59]: array([ 0, 18, 36, 54, 72, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 306, 324, 342]) Ok, it works, but: lionel[~]60>select( [t>90, t<300], [t,t] ) Sortie[60]: array([ 0, 18, 36, 54, 72, 90, 108, 126, 144, 162, 180, 198, 216, 234, 252, 270, 288, 306, 324, 342]) gives all items. How can I get out this problem? Thanks -- Lionel Roubeyrie - lroubeyrie(a)limair.asso.fr LIMAIR http://www.limair.asso.fr

1 0

Hamming low pass fir filter
by vincenzo cacciatore March 31, 2006

March 31, 2006

Hi all, i'm a new list member. I would like to design and apply a FIR low pass filter with Hamming window. I'm using scipy module. First I create the windows with the signal.firwin function and after i apply this window to my signal with the signal.lfilter function. I didn't understand where i can specify what kind of filter i'm implementing (low, high or band pass). Thanks all, Vincenzo

1 0

Re: [SciPy-user] SciPy-user Digest, Vol 31, Issue 42
by Pierre GM March 31, 2006

March 31, 2006

Salut, A couple of comments: > ##################################### > lionel[donn?es]21>from scipy import * > lionel[donn?es]22>from numpy import ma as MA > lionel[donn?es]23>test=MA.masked_object([1,2,3,4,nan,6], nan) > lionel[donn?es]24>print amin(test), amax(test), mean(test) It's a very bad idea to use nan as a masking value, as (nan == nan) is always False. The mask construction will fail, and you won't have any value actually masked. You can use `inf` instead, that seems to work. > lionel[donn?es]26>print amin(test), amax(test), mean(test) > 1.0 6.0 3.2 Try to use methods instead of functions. It should simplify your code and make it a bit more foolproof. A good part of scipy (a bit less of numpy) is not adapted to MaskedArrays. If you run into problems, please update the MaskedArray page of the wiki. A plus

2 1

Peak Fitting
by basvandijk＠home.nl March 31, 2006

March 31, 2006

Hello, I would like to integrate some data. The data has a somewhat similar form as the picture of the peak in [1]. I think a good function describing the data is: y = amp * exp( -exp (x - center/width ) - x - center/width + 1) Before I integrate I would like to fit it to get rid of the noise. I don't know much about linear algebra but I've read I can use a non linear least square fitting method to fit the data. Maybe I can also use others? My question is can I do this with Scipy and if so how? Greetings, Bas van Dijk [1] http://mathworld.wolfram.com/NonlinearLeastSquaresFitting.html

3 2