> > thank you for your answer > > actually i've to do some statistics (maximum,minimum,mean,standard > > deviation,....) of a file of data in which each column is a particular > > type of data. (the file is a tab separated value). > > I was trying to do this by using python (usually i work with fortran or > > bash, but i'm learning python), that the reason why i tried to use numpy.
numpy.loadtxt is probably what your looking for, provided your file contains only numbers and comment lines. check the example here: http://www.scipy.org/Numpy_Example_List_With_Doc#head-88ade192dacf0c15e4f1377096134ee559df07a0 X = numpy.loadtxt('data.txt') X[:,0].mean() # mean of first column X[:,0].max() # mux of first column X[:,0].var() # variance of first column ... enjoy! bernhard > > As I implied, you can do all this in standard Python "by hand", but > numpy/scipy can definitely make things easier. There's a dependency > cost here (your program needs one or more extra libraries to work), but > if that's no problem in your environment, I'd recommend that approach. > > The scipy add-on contains a bunch of things for file i/o; see > > http://www.scipy.org/doc/api_docs/SciPy.io.html > > for an overview. Since you're reading text files, the "array_import" > module seems to be what you need: > > http://www.scipy.org/doc/api_docs/SciPy.io.array_import.html > > There are active user communities for both numpy and scipy that can help > you sort out any remaining issues; for details, see: > > http://www.scipy.org/Mailing_Lists > > Hope this helps! > > </F> -- http://mail.python.org/mailman/listinfo/python-list
