[Numpy-discussion] array manupulation

[Olivier Delalleau]

Your array doesn't seem strange, it looks like a perfectly normal (11 x 5)
matrix of dtype float64.

>>> x = np.load('csum.npy')
>>> np.vstack((np.zeros((1, x.shape[1])), x))
array([[   0.        ,    0.        ,    0.        ,    0.        ,
0.        ],
       [  31.82571459,   29.0629995 ,   27.74400711,   26.6248159 ,
          25.73787976],
       [  59.82231014,   54.27656749,   51.87813602,   50.00937323,
          48.51771275],
       [  80.03460893,   73.46862838,   70.55710765,   68.412796  ,
          66.64323907],
       [  91.12613011,   85.96434025,   83.34633829,   81.36538282,
          79.70197141],
       [  96.11498624,   93.00049572,   91.13864656,   89.61535722,
          88.27247424],
       [  98.22403322,   96.55379518,   95.43277035,   94.39550817,
          93.42804   ],
       [  99.14200421,   98.27546395,   97.64792507,   97.00438205,
          96.3689249 ],
       [  99.55954577,   99.10418687,   98.76971791,   98.39724171,
          98.00386825],
       [  99.76081882,   99.51702755,   99.33960611,   99.13057243,
          98.9007987 ],
       [  99.8617198 ,   99.72882047,   99.63273748,   99.51539561,
          99.38460995],
       [ 100.        ,  100.        ,  100.        ,  100.        ,
100.        ]])

-=- Olivier


2013/5/26 Sudheer Joseph <sudheer.joseph at yahoo.com>

> Thank you Aronne for the helping hand,
>                                       I tried the transpose as a check
> when I could not get it correct other way. I could do it with test arrays,
> but it appears some thing strange happens when I do the cumsum. So I am
> attaching here the csum as csum.npy array, where I face problem if your
> time permits please see what happens with this strange array.!
>
>
> In [1]: csum=np.load('csum.npy') should get the array to you.
>
> This  array is obtained by doing a
> csum=np.cumsum(prcnt), which apparently doing some thing which I am not
> able to visualize.
>
> with best regards,
> Sudheer.
>
> >From:Aronne Merrelli <aronne.merrelli at gmail.com>
> >To:Discussion of Numerical Python <numpy-discussion at scipy.org>
> >Sent:Sunday, 26 May 2013 2:13 PM
> >Subject:Re: [Numpy-discussion] array manupulation
> >
> >
> >
> >
> >
> >On Sun, May 26, 2013 at 4:30 AM, Sudheer Joseph <sudheer.joseph at yahoo.com>
> wrote:
> >
> >Dear Brian,
> >>                I even tried below but no luck!
> >>In [138]: xx=np.zeros(11)
> >>In [139]: xx
> >>Out[139]: array([ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.])
> >>
> >>In [147]: xx.shape
> >>Out[147]: (11,)
> >>In [140]: xx=np.array(xx)[np.newaxis]
> >>In [141]: xx.shape
> >>Out[141]: (1, 11)
> >>In [142]: xx=xx.T
> >>In [143]: xx.shape
> >>Out[143]: (11, 1)
> >>In [144]: csum.shape
> >>Out[144]: (11, 5)
> >>In [145]: np.vstack((xx,csum))
> >>
>
> >>---------------------------------------------------------------------------
> >>ValueError                                Traceback (most recent call
> last)
> >>/media/SJOITB/SST_VAL/<ipython-input-145-2a0a60f68737> in <module>()
> >>----> 1 np.vstack((xx,csum))
> >>
> >>
> >>/usr/local/lib/python2.7/dist-packages/numpy-1.7.0-py2.7-linux-x86_64.egg/numpy/core/shape_base.pyc
> in vstack(tup)
> >>    224
> >>    225     """
> >>--> 226     return _nx.concatenate(map(atleast_2d,tup),0)
> >>    227
> >>    228 def hstack(tup):
> >>
> >>ValueError: all the input array dimensions except for the concatenation
> axis must match exactly
> >>
> >>
> >>
> >
> >
> >You've transposed the arrays, so now you need to stack the other way. So,
> you need to use hstack to concatenate arrays with the same column length
> (first axis), or vstack to concatenate arrays with the same row length
> (second axis). For example:
> >
> >
> >In [110]: xx1 = np.zeros((1,7)); cc1 = np.ones((3,7))
> >
> >
> >In [111]: xx2 = np.zeros((7,1)); cc2 = np.ones((7,3))
> >
> >
> >In [112]: np.vstack((xx1, cc1))
> >Out[112]:
> >array([[ 0.,  0.,  0.,  0.,  0.,  0.,  0.],
> >       [ 1.,  1.,  1.,  1.,  1.,  1.,  1.],
> >       [ 1.,  1.,  1.,  1.,  1.,  1.,  1.],
> >       [ 1.,  1.,  1.,  1.,  1.,  1.,  1.]])
> >
> >
> >In [113]: np.hstack((xx2, cc2))
> >Out[113]:
> >array([[ 0.,  1.,  1.,  1.],
> >       [ 0.,  1.,  1.,  1.],
> >       [ 0.,  1.,  1.,  1.],
> >       [ 0.,  1.,  1.,  1.],
> >       [ 0.,  1.,  1.,  1.],
> >       [ 0.,  1.,  1.,  1.],
> >       [ 0.,  1.,  1.,  1.]])
> >
> >
> >
> >
> >Also, I would highly recommend studying the NumPy for MATLAB users guide:
> >
> >
> >http://www.scipy.org/NumPy_for_Matlab_Users
> >
> >
> >
> >These issues (any many more) are discussed there.
> >
> >
> >
> >
> >Cheers,
> >Aronne
> >_______________________________________________
> >NumPy-Discussion mailing list
> >NumPy-Discussion at scipy.org
> >http://mail.scipy.org/mailman/listinfo/numpy-discussion
> >
> >
> >
>
> _______________________________________________
> NumPy-Discussion mailing list
> NumPy-Discussion at scipy.org
> http://mail.scipy.org/mailman/listinfo/numpy-discussion
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/numpy-discussion/attachments/20130526/bb5e4555/attachment.html>