In [55]:

    

import numpy


    

In [56]:

    

import scipy


    

In [57]:

    

import matplotlib


    

In [58]:

    

import sklearn


    

In [59]:

    

import psutil


    

In [60]:

    

import pandas


    

In [61]:

    

import IPython.parallel


    

    




/usr/local/lib/python2.7/dist-packages/IPython/parallel.py:13: ShimWarning: The `IPython.parallel` package has been deprecated. You should import from ipyparallel instead.
  "You should import from ipyparallel instead.", ShimWarning)

In [62]:

    

numpy.__path__


    

    
Out[62]:





['/usr/local/lib/python2.7/dist-packages/numpy']

In [63]:

    

numpy.__version__


    

    
Out[63]:





'1.10.1'

In [64]:

    

%run ../fetch_data.py


    

    




Using existing dataset folder:/home/janus/21_perspective/github/21v-python/unit_20/parallel_ml_tutorial-master/datasets
Checking availability of the 20 newsgroups dataset
Found archive: /home/janus/21_perspective/github/21v-python/unit_20/parallel_ml_tutorial-master/datasets/20news-bydate.tar.gz
Checking that the 20 newsgroups files exist...
=> Success!

In [1]:

    

import os
for fname in os.listdir('../datasets/'):
    print(fname)


    

    




20news-bydate.tar.gz
20news-bydate-train
20news-bydate-test

In [2]:

    

import numpy as np


    

In [3]:

    

a = np.array([1, 2, 3])


    

In [4]:

    

a


    

    
Out[4]:





array([1, 2, 3])

In [5]:

    

b = np.array([[0, 2, 4], [1, 3, 5]])


    

In [6]:

    

b


    

    
Out[6]:





array([[0, 2, 4],
       [1, 3, 5]])

In [7]:

    

b.shape


    

    
Out[7]:





(2, 3)

In [8]:

    

b.dtype


    

    
Out[8]:





dtype('int64')

In [9]:

    

a.shape


    

    
Out[9]:





(3,)

In [10]:

    

a.dtype


    

    
Out[10]:





dtype('int64')

In [11]:

    

np.zeros(5)


    

    
Out[11]:





array([ 0.,  0.,  0.,  0.,  0.])

In [12]:

    

np.ones(shape=(3, 4), dtype=np.int32)


    

    
Out[12]:





array([[1, 1, 1, 1],
       [1, 1, 1, 1],
       [1, 1, 1, 1]], dtype=int32)

In [13]:

    

c = b * 0.5


    

In [14]:

    

c


    

    
Out[14]:





array([[ 0. ,  1. ,  2. ],
       [ 0.5,  1.5,  2.5]])

In [15]:

    

c.shape


    

    
Out[15]:





(2, 3)

In [16]:

    

c.dtype


    

    
Out[16]:





dtype('float64')

In [17]:

    

a


    

    
Out[17]:





array([1, 2, 3])

In [18]:

    

d = a + c


    

In [19]:

    

d


    

    
Out[19]:





array([[ 1. ,  3. ,  5. ],
       [ 1.5,  3.5,  5.5]])

In [20]:

    

d[0]


    

    
Out[20]:





array([ 1.,  3.,  5.])

In [21]:

    

d[0, 0]


    

    
Out[21]:





1.0

In [22]:

    

d[:, 0]


    

    
Out[22]:





array([ 1. ,  1.5])

In [23]:

    

d.sum()


    

    
Out[23]:





19.5

In [24]:

    

d.mean()


    

    
Out[24]:





3.25

In [25]:

    

d.sum(axis=0)


    

    
Out[25]:





array([  2.5,   6.5,  10.5])

In [26]:

    

d.mean(axis=1)


    

    
Out[26]:





array([ 3. ,  3.5])

In [27]:

    

e = np.arange(12)


    

In [28]:

    

e


    

    
Out[28]:





array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11])

In [29]:

    

f = e.reshape(3, 4)


    

In [30]:

    

f


    

    
Out[30]:





array([[ 0,  1,  2,  3],
       [ 4,  5,  6,  7],
       [ 8,  9, 10, 11]])

In [31]:

    

e


    

    
Out[31]:





array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11])

In [32]:

    

e[5:] = 0


    

In [33]:

    

e


    

    
Out[33]:





array([0, 1, 2, 3, 4, 0, 0, 0, 0, 0, 0, 0])

In [34]:

    

f


    

    
Out[34]:





array([[0, 1, 2, 3],
       [4, 0, 0, 0],
       [0, 0, 0, 0]])

In [35]:

    

a


    

    
Out[35]:





array([1, 2, 3])

In [36]:

    

b


    

    
Out[36]:





array([[0, 2, 4],
       [1, 3, 5]])

In [37]:

    

d


    

    
Out[37]:





array([[ 1. ,  3. ,  5. ],
       [ 1.5,  3.5,  5.5]])

In [38]:

    

np.concatenate([a, a, a])


    

    
Out[38]:





array([1, 2, 3, 1, 2, 3, 1, 2, 3])

In [39]:

    

np.vstack([a, b, d])


    

    
Out[39]:





array([[ 1. ,  2. ,  3. ],
       [ 0. ,  2. ,  4. ],
       [ 1. ,  3. ,  5. ],
       [ 1. ,  3. ,  5. ],
       [ 1.5,  3.5,  5.5]])

In [40]:

    

np.hstack([b, d])


    

    
Out[40]:





array([[ 0. ,  2. ,  4. ,  1. ,  3. ,  5. ],
       [ 1. ,  3. ,  5. ,  1.5,  3.5,  5.5]])

In [41]:

    

%matplotlib inline


    

In [42]:

    

import matplotlib.pyplot as plt


    

In [43]:

    

x = np.linspace(0, 2, 10)


    

In [44]:

    

x


    

    
Out[44]:





array([ 0.        ,  0.22222222,  0.44444444,  0.66666667,  0.88888889,
        1.11111111,  1.33333333,  1.55555556,  1.77777778,  2.        ])

In [45]:

    

plt.plot(x, 'o-');


    

In [46]:

    

plt.plot(x, x, 'o-', label='linear')
plt.plot(x, x ** 2, 'x-', label='quadratic')

plt.legend(loc='best')
plt.title('Linear vs Quadratic progression')
plt.xlabel('Input')
plt.ylabel('Output');


    

In [47]:

    

samples = np.random.normal(loc=1.0, scale=0.5, size=1000)


    

In [48]:

    

samples.shape


    

    
Out[48]:





(1000,)

In [49]:

    

samples.dtype


    

    
Out[49]:





dtype('float64')

In [50]:

    

samples[:30]


    

    
Out[50]:





array([ 0.9723313 ,  1.05145447,  0.31986674, -0.26642655,  1.79545186,
        0.89663891, -0.78769912,  0.67095478,  1.28211239,  0.66237962,
        1.12488518,  1.76541164,  0.69160337,  0.05183348,  0.64934675,
        1.38717818,  0.37051531,  0.52509259,  1.17117495,  1.8224232 ,
        0.97875658,  1.60924242,  1.09170888,  0.40106831,  1.06174804,
        0.54542997,  1.21400516,  1.11310423, -0.1808943 , -0.36915236])

In [51]:

    

plt.hist(samples, bins=50);


    

In [52]:

    

samples_1 = np.random.normal(loc=1, scale=.5, size=10000)
samples_2 = np.random.standard_t(df=10, size=10000)


    

In [53]:

    

bins = np.linspace(-3, 3, 50)
_ = plt.hist(samples_1, bins=bins, alpha=0.5, label='samples 1')
_ = plt.hist(samples_2, bins=bins, alpha=0.5, label='samples 2')
plt.legend(loc='upper left');


    

In [54]:

    

plt.scatter(samples_1, samples_2, alpha=0.1);


    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]: