If `a` is your array:

    In [11]: a[:,:2]
    Out[11]: 
    array([[-0.57098887, -0.4274751 ],
           [-0.22279713, -0.51723555],
           [ 0.67492385, -0.69294472],
           [ 0.41086611,  0.26374238]])



I know this is quite an old question - 

    A = [[1, 2, 3],
         [4, 5, 6],
         [7, 8, 9]]

Let&#39;s say, you want to extract the first 2 rows and first 3 columns

    A_NEW = A[0:2, 0:3]
    A_NEW = [[1, 2, 3],
             [4, 5, 6]]

Understanding the syntax

    A_NEW = A[start_index_row : stop_index_row, 
              start_index_column : stop_index_column)]

If one wants row 2 and column 2 and 3

    A_NEW = A[1:2, 1:3]

Reference the numpy indexing and slicing article - [Indexing &amp; Slicing][1]


  [1]: http://docs.scipy.org/doc/numpy/reference/arrays.indexing.html

I have a `GLSurfaceView` set up and rendering as expected using a `GLSurfaceView.Renderer`. My App uses fragments from the android support package. When I navigate to a new fragment `surfaceDestroyed` is called but when I come back to the fragment via the backstack the `GLSurfaceView` will not render, calls to `requestRender` do not result in an `onDraw` call.

I am aware that I need to call `onResume` and `onPause` on the surface view and I am doing this from the hosting fragment but it doesn&#39;t seem to solve the issue. All examples about htis method refer to the activity, could this be the issue? And if so how do you use a `GLSurfaceView` inside a fragment.

Any insight greatly appreciated, I&#39;m happy to post code but it seems to be more of a general question to me,

Thanks

GLSurfaceView inside fragment not rendering when restarted

I recently installed Python 2.7.3 on a CentOS machine by compiling from source. Python 2.7.3 is installed at /opt/python2.7 and when I installed it I just changed /usr/bin/python to point to the new version. This apparently is wrong though because when I did it it broke yum. I would get the following.

    There was a problem importing one of the Python modules
    required to run yum. The error leading to this problem was:

       No module named yum

    Please install a package which provides this module, or
    verify that the module is installed correctly.

    It&#39;s possible that the above module doesn&#39;t match the
    current version of Python, which is:
    2.7.3 (default, May 15 2012, 17:45:42) 
    [GCC 4.4.4 20100726 (Red Hat 4.4.4-13)]

    If you cannot solve this problem yourself, please go to 
    the yum faq at:
      http://yum.baseurl.org/wiki/Faq


I changed /usr/bin/python to point back to the python 2.6.6 but now 2.6.6 is the default version of python. Any idea how to fix this?

Upgrade python without breaking yum


I have an array like this:

      array([[-0.57098887, -0.4274751 , -0.38459931, -0.58593526],
             [-0.22279713, -0.51723555,  0.82462029,  0.05319973],
             [ 0.67492385, -0.69294472, -0.2531966 ,  0.01403201],
             [ 0.41086611,  0.26374238,  0.32859738, -0.80848795]])


Now I want to extract the following:

       [-0.57098887, -0.4274751]
       [-0.22279713, -0.51723555]
       [ 0.67492385, -0.69294472]
       [ 0.41086611,  0.26374238]

So basically just first 2 columns..

Extracting first n columns of a numpy matrix

I have an array like this:
<pre><code class="hljs language-lua"> array([[-0.57098887, -0.4274751 , -0.38459931, -0.58593526],
 [-0.22279713, -0.51723555, 0.82462029, 0.05319973],
 [ 0.67492385, -0.69294472, -0.2531966 , 0.01403201],
 [ 0.41086611, 0.26374238, 0.32859738, -0.80848795]])
</code></pre>
Now I want to extract the following:
<pre><code class="hljs language-css"> [-0.57098887, -0.4274751]
 [-0.22279713, -0.51723555]
 [ 0.67492385, -0.69294472]
 [ 0.41086611, 0.26374238]
</code></pre>
So basically just first 2 columns..

I am looking for a python library for the Github APIv3 suitable for me.

I found one library [(python-github3)][1] mentioned in the [GH API docs][2]. After playing around with it in ipython for an hour or two, I found it really unintuitive to explore/work with. I looked some more, and found there&#39;s quite a number of people at least attempting to write such a library. The more promising-looking (at a glance) are [PyGithub][3] and [another python-github3][4], which apparently is different from the first one.

*Before I spend the next days consecutively trying out library after library, I wanted to ask the SO community if there is an accepted, definitive, obvious choice for that library?*

What I didn&#39;t like about the first library was the (to me) unintuitive way to get at data - some things you get as attributes, some you get as return value of a method, that return value is some complicated object which has to be paged and iterated through, etc. 

In that regard, PyGithub looks more attractive at first glance - clearly drill down through an object hierarchy, and then arrive at the attribute containing what you want:

`for repo in g.get_user().get_repos():
    print repo.name`

So, any pearls of wisdom to share? I know I don&#39;t have skills enough to quickly judge library quality, which is why I&#39;m turning to the SO community.

*edit: fwiw, I ended up using **PyGithub**. It works well, and the author is really receptive for feedback and bug reports. :-)*

  [1]: https://github.com/copitux/python-github3
  [2]: http://developer.github.com/v3/libraries/
  [3]: https://github.com/jacquev6/PyGithub
  [4]: https://github.com/ChristopherMacGown/python-github3

Most suitable python library for Github API v3

I am starting to depend heavily on the IPython notebook app to develop and document algorithms.  It is awesome; but there is something that seems like it should be possible, but I can&#39;t figure out how to do it:

I would like to insert a local image into my (local) IPython notebook markdown to aid in documenting an algorithm.  I know enough to add something like `&lt;img src=&quot;image.png&quot;&gt;` to the markdown, but that is about as far as my knowledge goes.  I assume I could put the image in the directory represented by 127.0.0.1:8888 (or some subdirectory) to be able to access it, but I can&#39;t figure out where that directory is.  (I&#39;m working on a mac.)  So, is it possible to do what I&#39;m trying to do without too much trouble?


Inserting image into IPython notebook markdown

I am working on **selenium automation project** using Python.

I am facing an issue, which is handling multiple browser windows.

Scenario is as follows. When I click a link on the home page, a new window opens. In the newly opened window I cannot perform any actions, because the focus is still on the home page web driver.

Can anybody show me how to change focus from the background window to the newly opened window?

A possible solution is `driver.switch_to.window()`, but it requires the window&#39;s name. How to find out the window&#39;s name? If this is a wrong way to do this, can anybody give some code examples to perform this action?

How to switch to new window in Selenium for Python?

I would like to read some characters from a string `s1` and put it into another string `s2`.

However, assigning to `s2[j]` gives an error:

```python
s2[j] = s1[i]

# TypeError: &#39;str&#39; object does not support item assignment
```

---


In C, this works:

```cpp
int i = j = 0;
while (s1[i] != &#39;\0&#39;)
    s2[j++] = s1[i++];
```

My attempt in Python:

```python
s1 = &quot;Hello World&quot;
s2 = &quot;&quot;
j = 0

for i in range(len(s1)):
    s2[j] = s1[i]
    j = j + 1
```

&#39;str&#39; object does not support item assignment

With Sqlite, a `select .. from` command returns the results `output`, which prints:

    &gt;&gt;print output
    [(12.2817, 12.2817), (0, 0), (8.52, 8.52)]

It seems to be a list of tuples. I would like to either convert `output` to a simple list:

    [12.2817, 12.2817, 0, 0, 8.52, 8.52]

or a 2x3 matrix:

    12.2817 12.2817
    0          0 
    8.52     8.52

to be read via `output[i][j]`

The flatten command does not do the job for the 1st option, and I have no idea for the second one...

A fast solution would be appreciated, as the real data is much bigger.



Transform &quot;list of tuples&quot; into a flat list or a matrix

I load some machine learning data from a CSV file. The first 2 columns are observations and the remaining columns are features.

Currently, I do the following:

    data = pandas.read_csv(&#39;mydata.csv&#39;)

which gives something like:

    data = pandas.DataFrame(np.random.rand(10,5), columns = list(&#39;abcde&#39;))

I&#39;d like to slice this dataframe in two dataframes: one containing the columns `a` and `b` and one containing the columns `c`, `d` and `e`.

It is not possible to write something like 

    observations = data[:&#39;c&#39;]
    features = data[&#39;c&#39;:]

I&#39;m not sure what the best method is. Do I need a `pd.Panel`?

By the way, I find dataframe indexing pretty inconsistent: `data[&#39;a&#39;]` is permitted, but `data[0]` is not. On the other side, `data[&#39;a&#39;:]` is not permitted but `data[0:]` is.
Is there a practical reason for this? This is really confusing if columns are indexed by Int, given that `data[0] != data[0:1]`

How to take column-slices of dataframe in pandas

When iterating over a large array with a range expression, should I use Python&#39;s built-in `range` function, or numpy&#39;s `arange` to get the best performance?

My reasoning so far: 

`range` probably resorts to a native implementation and might be faster therefore. On the other hand, `arange` returns a full array, which occupies memory, so there might be an overhead. Python 3&#39;s range expression is a generator, which does not hold all the values in memory.

built-in range or numpy.arange: which is more efficient?

For my unittest, I want to check if two arrays are identical.  Reduced example: 

    a = np.array([1, 2, np.NaN])
    b = np.array([1, 2, np.NaN])

    if np.all(a==b):
        print &#39;arrays are equal&#39;

This does not work because `nan != nan`.
What is the best way to proceed?

Comparing numpy arrays containing NaN

I have a `DataFrame` with a few columns. One columns contains a symbol for which currency is being used, for instance a euro or a dollar sign. Another column contains a budget value. So for instance in one row it could mean a budget of 5000 in euro and in the next row it could say a budget of 2000 in dollar.

In pandas I would like to add an extra column to my DataFrame, normalizing the budgets in euro. So basically, for each row the value in the new column should be the value from the budget column * 1 if the symbol in the currency column is a euro sign, and the value in the new column should be the value of the budget column * 0.78125 if the symbol in the currency column is a dollar sign.

I know how to add a column, fill it with values, copy values from another column etc. but not how to fill the new column conditionally based on the value of another column. 

Any suggestions?

Conditionally fill column values based on another columns value in pandas

Suppose I have a large in memory numpy array, I have a function `func` that takes in this giant array as input (together with some other parameters). `func` with different parameters can be run in parallel. For example:

    def func(arr, param):
        # do stuff to arr, param
    
    # build array arr
    
    pool = Pool(processes = 6)
    results = [pool.apply_async(func, [arr, param]) for param in all_params]
    output = [res.get() for res in results]
        
If I use multiprocessing library, then that giant array will be copied for multiple times into different processes. 

Is there a way to let different processes share the same array? This array object is read-only and will never be modified. 

What&#39;s more complicated, if arr is not an array, but an arbitrary python object, is there a way to share it? 

[EDITED]

I read the answer but I am still a bit confused. Since fork() is copy-on-write, we should not invoke any additional cost when spawning new processes in python multiprocessing library. But the following code suggests there is a huge overhead: 

    from multiprocessing import Pool, Manager
    import numpy as np; 
    import time
    
    def f(arr):
        return len(arr)
    
    t = time.time()
    arr = np.arange(10000000)
    print &quot;construct array = &quot;, time.time() - t;
    
    
    pool = Pool(processes = 6)
    
    t = time.time()
    res = pool.apply_async(f, [arr,])
    res.get()
    print &quot;multiprocessing overhead = &quot;, time.time() - t;

output (and by the way, the cost increases as the size of the array increases, so I suspect there is still overhead related to memory copying): 

&lt;!-- language: lang-none --&gt;

    construct array =  0.0178790092468
    multiprocessing overhead =  0.252444982529

Why is there such huge overhead, if we didn&#39;t copy the array? And what part does the shared memory save me? 


Content Type	Original Author	Original Content on Stackoverflow
Question	frazman	View Question on Stackoverflow
Solution 1 - Python	NPE	View Answer on Stackoverflow
Solution 2 - Python	yashgarg1232	View Answer on Stackoverflow

Extracting first n columns of a numpy matrix

Python Problem Overview

Python Solutions

Solution 1 - Python

Solution 2 - Python

Upgrade python without breaking yum

GLSurfaceView inside fragment not rendering when restarted

Attributions