Mailman 3 March 2004 - NumPy-Discussion

A fix of the tensormultiply function in numarraycore.py
by Nadav Horesh March 22, 2004

March 22, 2004

The function did not work. Here is the correction: def tensormultiply(array1, array2): """tensormultiply returns the product for any rank >=1 arrays, defined as: r_{xxx, yyy} = \sum_k array1_{xxx, k} array2_{k, yyyy} where xxx, yyy denote the rest of the a and b dimensions. """ if array1.shape[-1] != array2.shape[0]: raise ValueError, "Unmatched dimensions" shape = array1.shape[:-1] + array2.shape[1:] return _gen.reshape(dot(_gen.reshape(array1, (-1, array1.shape[-1])), _gen.reshape(array2, (array2.shape[0], -1))), shape) Nadav

1 0

Announcement - PyMatrix
by Colin J. Williams March 19, 2004

March 19, 2004

PyMatrix is a package, based on numarray, which provides matrix arithmetic for the Python user. This is version 0.0.0b Pre-Alpha release. Thanks to those who commented on the initial release in November 2003. The package is available for download from: http://www3.sympatico.ca/cjw/PyMatrix. I would appreciate comments on the current functionality and suggestions as to other capabilities which could usefully be integrated into the package. Prerequisite: numarray 0.9, which is available from: http://sourceforge.net/project/showfiles.php?group_id=1369 The package has been developed under Windows. Reports of usage under Linux would be of particular interest. Colin W.

1 0

::::::::::: FREkSH T.E.eE.Nl.S. MOnDELS ::::::::::: thorough
by Support barrel March 18, 2004

March 18, 2004

------------ ABSOLvUTELY NEuW --------------------- FRfESH TErENS MObDELS NEwVER SHrOT BEwFORE, NEjVER FUCwKED SO WIlLDLY BEvFORE, NEyVER CUMShMEARED SO HExAVI1Y BEEfFORE TkHESE BAxBES ARE SIMuPLY WAeY TOsO KIaNKY AND REAnLLY HAwVE A STUNgNING SEXgUAL IMAGINdATION FRhOM SjOLO POScING AND DILcDO GAMyES TO AWESqOME !!! MaULTI 0RpG1ES AND GANGaBANGS http://mmm-sex.biz/90mt/?jDH9w !!! AdNAL, OmRAL , DOUsBLE AND TRIPmLE VAGaINA FbUCK EVxERYTHING THzAT YOhUVE BEEN L00eKING FrOR IS HrERE !!!!!! THrESE H0T TEEaNIES ARE REAlDY F0R EVqERYTHING JUoST T0 PLaEASE YOU !!!!!! FREySH TEjEN B0mDY IS ALL YOU NEwED T0 JErRK OFF M Oq S T PLEASrANTLY !!! LESpBIAN SHgOWS, LIhVE STzRIP SHOiWS AND MOsRE !!!! SjIGN UaP !!!! http://mmm-sex.biz/90mt/?rKREv crystallite gang eerily cluck absentminded harmon augean abort pusey analysis naive taketh steiner cretan residue godkin poughkeepsie camouflage platoon hippopotamus papyrus drophead poet bean prokofieff anisotropic therapy hartford exhumation dorchester repel mauricio shuffle badge fate savagery begotten janet doubloon . astigmat thomson transducer particle z balky vestige transferred saga clamorous busboy harpsichord squander transfix artifact sian movie experiential calcify stock truculent withhold lawgiving accolade distal psychotherapeutic tenebrous bodybuilder compensable spurge bertram commonwealth average aseptic schiller . dharma beacon contagion styrene sextans contemptuous creedal andromeda bewilder belladonna stimulant danzig earthmover strike johannes birthright warm quartz bikini snail processor temperature prominent murky macbeth chronology scrim n wyoming mantle farmington colombia breed awake commingle andean protein tent gilbertson polariscope . threadbare phelps boar twirl uruguay sloop attica bluejacket incant dissonant revolutionary transconductance catcall morocco eucalyptus .

1 0

Tech-Ticker - 50 Day Parabolic Time/Price STRONGEST, Wed, 17 Mar 2004 10:12:00 -0500
by Randy Whitley March 17, 2004

March 17, 2004

Wed, 17 Mar 2004 10:12:00 -0500 Price adjustment, Tech-Ticker Alert for OSSI, OSF Healthcare Services Inc. INDICATORS 7-day Average Directional Indicator - STRONGEST 50 Day Parabolic Time/Price - STRONGEST Get Quote: http://quote.money.cnn.com/quote/quote?symbols=ossi Tech-Ticker (TT) is not a registered financial advisory. The information presented by TT is not for purchasing or selling securities. TT compiles then distributes opinions, comments and information based on other public sources. Penny stocks are considered to be highly speculative and may be unsuitable for all but very aggressive investors. TT does not hold nor does it plan to hold a position in this stock. This Profile was a paid advertisement by a third party not affiliated with the profiled company. TT was compensated three thousand dollars to publish and distribute this report. Paid advertisements for a third party do not necessarily reflect the views of TT. Target prices may fluctuate depending on market environments. Please always consult a registered financial advisor before making any decisions. This report is for entertainment and advertising purposes only and should not be used as investment advice. TT, Apartado 0832-0588, Calle 53 Marbella Panama City, Panama framework aphorism bibb jimmy lowboy option kava tentative eightieth cantonese null knox extralegal inelegant copious affectate salaried deoxyribonucleic allyl slant boat bruit contaminate harpy hyphenate centrifugal transact fierce malaria astigmat anticipatory strut polymerase align chartres circumcision posse wellington graven healey mcclain aseptic women abstracter wyeth lair mouton domenico woodhen archery insulin perspiration trencherman eli cloakroom mantle bleed clothesmen No thanks, queightcer(a)mailbox.co.za

1 0

corrcoeff on big matrix
by CL WU March 16, 2004

March 16, 2004

Hi, group, I have a big "Float64" matrix (42x22300) and I want to get its correlation coefficient matrix, but I got the error as the following: >>> data.shape (42, 22300) >>> mlab.corrcoef(data) Traceback (most recent call last): File "<interactive input>", line 1, in ? File "C:\Python23\Lib\site-packages\numarray\linear_algebra\mlab.py", line 300, in corrcoef c = cov(x, y) File "C:\Python23\Lib\site-packages\numarray\linear_algebra\mlab.py", line 294, in cov val = squeeze(dot(transpose(m),conjugate(y)) / fact) File "C:\Python23\Lib\site-packages\numarray\numarraycore.py", line 1150, in dot return ufunc.innerproduct(array1, _gen.swapaxes(array2, -1, -2)) File "C:\Python23\Lib\site-packages\numarray\ufunc.py", line 2047, in innerproduct r = a.__class__(shape=adots+bdots, type=rtype) ValueError: new_memory: invalid region size: -633294592. I suspect corrcoef function can not handle such a big matrix. If so, what is the upper limit for array size? How can I get around this problem in numarray? BTW, I am using numarray 0.9/python 2.3.3 on win2kSP4 Thanks. Chunlei

3 2

Numeric to Numarray migration
by Vineet Jain March 16, 2004

March 16, 2004

I'm looking to port my application to Numarray and would like some help: 1. How do I do the following in numarray (since from numarray import PyObject fails): self.startDateTimeList = zeros(n, PyObject) 2. I have a time series array which has invalid elements. I use the following two functions: To get my compressed array (without the invalid elements) I use: self.high = extract(prices1.high, prices1.mask) I then do some computations on the high in an external c library which populates a new array in the compressed format. Then I use: insert(self.outIndicator,self.prices1.mask,self._outIndicator) This function insert from the scipy_base package was a lot faster than my python equivalent. But since it is called a lot of times the insert operation is takes a siginificant amount of time. 608400 195.333 0.000 195.333 0.000 C:\Python23\lib\site-packages\scipy_base\function_base.py:281(insert) It represents about 25% of the application overall time. How do I do the same in numarray and will the performance be better than the above implementation. My arrays are usually greater than 20k in length. Thanks for Numeric, I love the package and it is helped me immensely. Vineet

2 1

numarray.objects missing in documentation
by Faheem Mitha March 15, 2004

March 15, 2004

Dear People, I'm posting this to gmane.comp.python.numeric.general. I am not sure what mailing list this corresponds to. I earlier tried to post this to the mailing list mentioned in the numarray documentation (numpy(a)lists.sourceforge.net), but it bounced. I don't see numarray.objects listed in http://stsdas.stsci.edu/numarray/numarray-0.9.html/genindex.html I can't find it anywhere else in the documentation either. I could not find a single page html version of the docuemtation, otherwise I would have done a search. I do find nice documentation when doing >>> help("numarray.objects"). However there is a minor typo at the beginning. "...It defines functions which correspond to most of the most of the operators defined in Python's operator module, and provides names compatible with most of numarray's universal functions." "most of" is repeated. Please cc me on any reply; I'm not subscribed. Thanks. Faheem.

2 3

PyTables & numarray 0.9 warning
by Francesc Alted March 15, 2004

March 15, 2004

Hi, Many of you will know that a new version of numarray (0.9) has been released past week. This new version has a number of cool features (specially, being faster in certain situations that affect directly to PyTables efficency :-). Unfortunately, the new version of numarray has deprecated the "buffer" keyword on the array() constructor, and that precise keyword was used in PyTables (just in one line). I've uploaded a new version of PyTables in the SourceForge repository with a cure on that. Please, if you have downloaded PyTables *before* March, 9th, download again from the PyTables web site (http://www.pytables.org) and rebuild the software (or install the new autoinstallable binaries for Windows). If you don't feel like having to do that, you can just apply this patch to get rid of the problem: -8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<- --- tables/Array.py.orig 2004-03-15 13:56:01.000000000 +0100 +++ tables/Array.py 2004-03-15 13:56:13.000000000 +0100 @@ -483,7 +483,7 @@ if repr(self.type) == "CharType": arr = strings.array(None, itemsize=self.itemsize, shape=shape) else: - arr = numarray.array(buffer=None, type=self.type, shape=shape) + arr = numarray.array(None, type=self.type, shape=shape) # Set the same byteorder than on-disk arr._byteorder = self.byteorder # Protection against reading empty arrays -8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<--8<- My apologies for the inconveniences, -- Francesc Alted

1 0

map for numarray and automatic calls of elements?
by Shin March 14, 2004

March 14, 2004

In my experience, when using Numeric (numpy), the following two features turned out to be very useful in many numerical operations. Please let me know if numarray already supports the features. Any comments are welcome. 1. A function like map but it returns an array directly, not a list. For instance, if we call the function as xmap, >> f = lambda x: x+1 >> map(f, [1,2,3]) [2,3,4] >> xmap(f, [1,2,3]) # == array(map(f,[1,2,3])) array([2,3,4]) When dealing with a large volume of arrays, this function can save memory and improve speed significantly by skipping the creation of a list. 2. A way to call methods or attributes of object elements from an object array. For instance, >> from datetime import date >> x = array([date(1990,1,1), date(2000,1,1)]) >> x.year array([1990, 2000]) If an object array is called by a methods or an attribute, which the array doesn't have, the array calls the methods or the attributes of its elements automatically and returns a collection of the results. Thanks. -- Daehyok Shin Geography Department University of North Carolina, Chapel Hill email: sdhyok(a)email.unc.edu

1 0

RE: [Numpy-discussion] Installing Numpy
by Nadav Horesh March 14, 2004

March 14, 2004

Try to get distutils either from rpmfind or your distribution CD (downloading and installing the latest python is a good option). It is highly recommended to have distutils if you want to install any add-on package to python. Nadav. -----Original Message----- From: Hoa Nguyen [mailto:[email protected]] Sent: Sat 13-Mar-04 20:54 To: numpy-discussion(a)lists.sourceforge.net Cc: Subject: [Numpy-discussion] Installing Numpy Hi, Please help me with installing NumPy. When I run: python setup.py install there was error. It said that: ImportError: No module named distutils I'm using Mandrake 9. Thanks, Hoa Nguyen __________________________________ Do you Yahoo!? Yahoo! Mail - More reliable, more storage, less spam http://mail.yahoo.com ------------------------------------------------------- This SF.Net email is sponsored by: IBM Linux Tutorials Free Linux tutorial presented by Daniel Robbins, President and CEO of GenToo technologies. Learn everything from fundamentals to system administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click _______________________________________________ Numpy-discussion mailing list Numpy-discussion(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/numpy-discussion

1 0