Mailman 3 September 2010 - NumPy-Discussion

Re: [Numpy-discussion] [NumPy-Tickets] [NumPy] #1348: CHM of Numpy Reference Guide (development version) is outdated
by Sebastian Haase 06 Sep '10

06 Sep '10

Hi, is there an URL of the weekly built CHM documentation file ? Thanks, Sebastian Haase On Sun, Sep 5, 2010 at 6:39 PM, <numpy-tickets(a)scipy.org> wrote: > #1348: CHM of Numpy Reference Guide (development version) is outdated > ----------------------------+----------------------------------------------- > Reporter: petar | Owner: pv > Type: defect | Status: closed > Priority: normal | Milestone: > Component: Documentation | Version: > Resolution: fixed | Keywords: > ----------------------------+----------------------------------------------- > Changes (by pv): > > * status: new => closed > * resolution: => fixed > > > Comment: > > You need to have installed http://sphinx.pocoo.org/ version >= 1.0.1, > Scipy, and matplotlib. > > Anyway, I set up scripts for rebuilding the CHM weekly, so from now on it > should stay up to date. >

2 2

PyGTK breaks fortran extension???
by Åsmund Hjulstad 06 Sep '10

06 Sep '10

Hi, all, I have a f2py wrapped fortran extension, compiled using gcc-mingw32 (v.4.5.0), numpy 1.5, Python 2.7, where I am experiencing the strangest behaviour. It appears that loading pygtk breaks my fortran extension. The fortran code has initialization code (it calculates reference state for a combination of fluids). The initialization appearantly has some sanity checks, verifying the results. This initialization fails if gtk is loaded first. import refprop from gtk import _gtk # if this line is commented out the next line works as expected ierr,herr = refprop.setup0(2,"METHANE.FLD|ETHANE.FLD","HMX.BNC","DEF") I initially started suspecting matplotlib, but as the problem only appeared when loading a gtk backend, I narrowed it down to this particular import. The following imports break it in just the same way (with, I suspect, the gtk import still being the culprit). import matplotlib matplotlib.use('GTK') import matplotlib.pyplot I have little idea how to debug this further, and probably will just choose to use a different backend. Has anybody else encountered something like this? Is is possible that msvcrt.dll has something to do with this? I haven't been able to get my library to link to msvcr90, I guess I have to do something with a manifest, but I haven't grasped it yet. Before using the gtk backend of matplotlib, everything worked OK, so I thought it wasn't a problem. I have a VS2008PRO DVD on the way in the mail, but was hoping to be able to fix this with mingw. Best regards, Åsmund Hjulstad, asmund(a)hjulstad.com

2 2

test if two arrays share the same data
by Ernest Adrogué 05 Sep '10

05 Sep '10

Hi, How can it be done? id() doesn't do it: In [238]: a= np.arange(5) In [239]: id(a) == id(a[:]) Out[239]: False Any ideas? Ernest

5 6

Add function for creating recarray from database query?
by John Salvatier 04 Sep '10

04 Sep '10

Hello, I recently had to get data from a mysql database into a recarray. The result was not very long but nontrivial to figure out: def recarray_from_db(db, command): """ executes a command and turns the results into a numpy recarray (record array)""" cursor = db.cursor() cursor.execute(command) column_dtypes = [(col[0], _convert(col[1])) for col in cursor.description] return np.fromiter((tuple (row) for row in cursor), dtype=column_dtypes, count = cursor.rowcount) _type_conversions = {decimal.Decimal : float} def _convert(type): try : return _type_conversions[type] except: return type It uses only the Python database API. Would something like this be a useful addition to numpy (maybe have it take a cursor object instead of a connection object)? I also found the following function useful for results with timedate columns, since numpy does not yet have a datetime dtype. def time_column(date, column_name): """ makes a string for calculating a time as (decimal) number of days since a date for a MySQL column. This is because the numpy.datetime datatype is not well developed.""" return "TIME_TO_SEC(timediff(" + column_name + ",'" + str(datetime.strptime(date, date_format)) + "'))/(60*60*24) as " + column_name date_format = '%Y-%m-%d' Best Regards, John

2 1

Cutoff small values of a vector to stop underflow
by Rick Muller 03 Sep '10

03 Sep '10

There just *has* to be a better way of doing this. I want to cut off small values of a vector, and I'm currently doing something like: >>> for i in xrange(n): >>> if abs(A[i]) < tol: A[i] = 0 Which is slow, since A can be really long. Is there a way to write a Ufunc that would do something like this more directly? The naive things I've tried didn't work, e.g. >>> def cut(a,tol=1e-10): >>> if less(absolute(a),tol): >>> a = 0 >>> return -- Rick Muller rpmuller(a)gmail.com 505-750-7557

2 2

recarray to csv
by Guillaume Chérel 03 Sep '10

03 Sep '10

Hello, I'd like to know if there is a convenient routine to write recarrays into cvs files, with the first line of the file being the name of the fields. Thanks, Guillaume

5 6

Simplified question on tensordot
by Rick Muller 03 Sep '10

03 Sep '10

Sorry for the rapid repost, but I thought of a much easier way to ask the question I asked a few minutes ago. I have two matrices, A and B, both of which are n x m. n is big (~10,000), and m is small (~10). I want to take the product AB such that I get a length-n vector, as in: >>> AB = zeros(n,'d') >>> for i in xrange(n): >>> AB[i] = dot(A[i,:],B[i,:]) only I don't want the for-loop. The various things I've tried with dot and tensordot inevitably give me a n x n matrix, which isn't what I want. Can anyone help me find an easier way to do this? -- Rick Muller rpmuller(a)gmail.com

3 3

Histogram does not preserve subclasses of ndarray (e.g. masked arrays)
by Joe Kington 02 Sep '10

02 Sep '10

Hi all, I just wanted to check if this would be considered a bug. numpy.histogram does not appear to preserve subclasses of ndarrays (e.g. masked arrays). This leads to considerable problems when working with masked arrays. (As per this Stack Overflow question<http://stackoverflow.com/questions/3610040/how-to-create-the-histogram-of-a…> ) E.g. import numpy as np x = np.arange(100) x = np.ma.masked_where(x > 30, x) counts, bin_edges = np.histogram(x) yields: counts --> array([10, 10, 10, 10, 10, 10, 10, 10, 10, 10]) bin_edges --> array([ 0. , 9.9, 19.8, 29.7, 39.6, 49.5, 59.4, 69.3, 79.2, 89.1, 99. ]) I would have expected histogram to ignore the masked portion of the data. Is this a bug, or expected behavior? I'll open a bug report, if it's not expected behavior... This would appear to be easily fixed by using asanyarray rather than asarray within histogram. E.g. this diff for numpy/lib/function_base.py Index: function_base.py =================================================================== --- function_base.py (revision 8604) +++ function_base.py (working copy) @@ -132,9 +132,9 @@ """ - a = asarray(a) + a = asanyarray(a) if weights is not None: - weights = asarray(weights) + weights = asanyarray(weights) if np.any(weights.shape != a.shape): raise ValueError( 'weights should have the same shape as a.') @@ -156,7 +156,7 @@ mx += 0.5 bins = linspace(mn, mx, bins+1, endpoint=True) else: - bins = asarray(bins) + bins = asanyarray(bins) if (np.diff(bins) < 0).any(): raise AttributeError( 'bins must increase monotonically.') Thanks! -Joe

3 3

Boolean arrays
by Brett Olsen 02 Sep '10

02 Sep '10

Hello, I have an array of non-numeric data, and I want to create a boolean array denoting whether each element in this array is a "valid" value or not. This is straightforward if there's only one possible valid value: >>> import numpy as N >>> ar = N.array(("a", "b", "c", "b", "b", "a", "d", "c", "a")) >>> ar == "a" array([ True, False, False, False, False, True, False, False, True], dtype=bool) If there's multiple possible valid values, I've come up with a couple possible methods, but they all seem to be inefficient or kludges: >>> valid = N.array(("a", "c")) >>> (ar == valid[0]) | (ar == valid[1]) array([ True, False, True, False, False, True, False, True, True], dtype=bool) >>> N.array(map(lambda x: x in valid, ar)) array([ True, False, True, False, False, True, False, True, True], dtype=bool) Is there a numpy-appropriate way to do this? Thanks, Brett Olsen

9 11

Github migration?
by Travis Oliphant 02 Sep '10

02 Sep '10

Hi all, I'm curious as to the status of the Github migration and if there is anything I can do to help. I have a couple of weeks right now and I would love to see us make the transition of both NumPy and SciPy to GIT. On a slightly related note, it would really help the numpy-refactor project if it received more input from others in the community. Right now, the numpy-refactor is happening on a public github branch (named numpy-refactor) awaiting numpy itself to be on github. It would be more useful if the numpy-refactor branch were a regular branch of the github NumPy project. The numpy-refactor project is making great progress and we have a working core library that can be built on Windows, Mac, and Linux. The goal is for this numpy-refactor to become the basis for NumPy 2.0 which should come out sometime this Fall. Already, a lot of unit tests have been written and code coverage has increased on the core NumPy code which I think we all agree is "a good thing" In addition, some lingering reference count bugs (particularly in object arrays) have been found and squashed. There is also some good progress on the Cython backend for .NET which would allow and also put pressure on migration of most of SciPy to Cython-or-Fwrap generated extension modules. I am looking forward to working on NumPy a little more over the coming months. All the best, -Travis

6 23