In [1]:import numpy as np


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

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

In [3]:a[1]

Out[3]:2

In [4]:a.shape

Out[4]:(3,)

In [5]:a.dtype.name

Out[5]:'int64'

In [6]:b = np.arange(4) # 0から3（=4の一つ手前）まで
b

Out[6]:array([0, 1, 2, 3])

In [7]:c = np.arange(2, 4) # 2から3（=4の一つ手前）まで
c

Out[7]:array([2, 3])

In [8]:d = np.arange(3, 4, 0.1) # 3から0.1飛ばしで4の直前まで
d

Out[8]:array([ 3. ,  3.1,  3.2,  3.3,  3.4,  3.5,  3.6,  3.7,  3.8,  3.9])

In [9]:a = np.array([1, 2, 3])
a.dtype.name

Out[9]:'int64'

In [10]:b = np.array([1., 2., 3.])
b.dtype.name

Out[10]:'float64'

In [11]:c = np.array([1, 2, 3], dtype=np.float64)
c.dtype.name

Out[11]:'float64'


In [12]:l = [[1, 2, 3], [4, 5, 6]]
a = np.array(l)
a

Out[12]:array([[1, 2, 3],
       [4, 5, 6]])

In [13]:a[1, 2]

Out[13]:6


In [14]:!echo "1,2,3\n4,5,6" >hoge.csv # まずはテスト用のデータを作る

In [15]:import csv

def read_csv_as_array(filename):
    data = []
    with open(filename) as fp:
        for row in csv.reader(fp):
            data.append([float(x) for x in row])
    return np.array(data)

data = read_csv_as_array("hoge.csv")
data

Out[15]:array([[ 1.,  2.,  3.],
       [ 4.,  5.,  6.]])

In [16]:a = np.zeros((3, 4))
a

Out[16]:array([[ 0.,  0.,  0.,  0.],
       [ 0.,  0.,  0.,  0.],
       [ 0.,  0.,  0.,  0.]])

In [17]:b = np.ones((4, 4))
b

Out[17]:array([[ 1.,  1.,  1.,  1.],
       [ 1.,  1.,  1.,  1.],
       [ 1.,  1.,  1.,  1.],
       [ 1.,  1.,  1.,  1.]])

In [18]:c = np.zeros((3, 3), dtype=np.int64)
c

Out[18]:array([[0, 0, 0],
       [0, 0, 0],
       [0, 0, 0]])

In [19]:e = np.empty((4, 3))
e

Out[19]:array([[ 0.,  0.,  0.],
       [ 0.,  0.,  0.],
       [ 0.,  0.,  0.],
       [ 0.,  0.,  0.]])

In [20]:np.random.seed(0)
a = np.random.rand(3, 4)
a

Out[20]:array([[ 0.5488135 ,  0.71518937,  0.60276338,  0.54488318],
       [ 0.4236548 ,  0.64589411,  0.43758721,  0.891773  ],
       [ 0.96366276,  0.38344152,  0.79172504,  0.52889492]])

In [21]:b = np.random.randn(3, 4)
b

Out[21]:array([[ 0.76103773,  0.12167502,  0.44386323,  0.33367433],
       [ 1.49407907, -0.20515826,  0.3130677 , -0.85409574],
       [-2.55298982,  0.6536186 ,  0.8644362 , -0.74216502]])


In [22]:v = np.array([1, 2, 3, 4])
v[1:3]

Out[22]:array([2, 3])

In [23]:v[2:]

Out[23]:array([3, 4])

In [24]:v[:3]

Out[24]:array([1, 2, 3])

In [25]:v[:-1]

Out[25]:array([1, 2, 3])

In [26]:a = np.arange(9).reshape(3, 3)
a

Out[26]:array([[0, 1, 2],
       [3, 4, 5],
       [6, 7, 8]])

In [27]:a[:2, :]

Out[27]:array([[0, 1, 2],
       [3, 4, 5]])

In [28]:a[:, 1:]

Out[28]:array([[1, 2],
       [4, 5],
       [7, 8]])

In [29]:a[:, 2]

Out[29]:array([2, 5, 8])

In [30]:a = np.array([5, 4, 3, 2, 1])
ii = np.array([0, 2, 3])
a[ii]

Out[30]:array([5, 3, 2])

In [31]:b = np.arange(9).reshape(3, 3)
b

Out[31]:array([[0, 1, 2],
       [3, 4, 5],
       [6, 7, 8]])

In [32]:ii = np.array([0, 1])
jj = np.array([1, 2])
b[ii, jj]

Out[32]:array([1, 5])

In [33]:ii = np.array([True, False, True, False, False])
a[ii]  # a.shape == ii.shape

Out[33]:array([5, 3])

In [34]:kk = np.array([[True, False, False], [False, True, False], [False, False, False]])
b[kk]  # b.shape == kk.shape

Out[34]:array([0, 4])

In [35]:a = np.arange(20).reshape(5, 4)
a

Out[35]:array([[ 0,  1,  2,  3],
       [ 4,  5,  6,  7],
       [ 8,  9, 10, 11],
       [12, 13, 14, 15],
       [16, 17, 18, 19]])

In [36]:a.sum()

Out[36]:190

In [37]:a.sum(axis=0)

Out[37]:array([40, 45, 50, 55])

In [38]:a.mean(axis=1)

Out[38]:array([  1.5,   5.5,   9.5,  13.5,  17.5])

In [39]:a.max(axis=1)

Out[39]:array([ 3,  7, 11, 15, 19])

In [40]:a.min(axis=1)

Out[40]:array([ 0,  4,  8, 12, 16])

In [41]:a.argmax(axis=1)

Out[41]:array([3, 3, 3, 3, 3])

In [42]:a.argmin(axis=1)

Out[42]:array([0, 0, 0, 0, 0])


In [43]:a = np.arange(9).reshape(3, 3)
a

Out[43]:array([[0, 1, 2],
       [3, 4, 5],
       [6, 7, 8]])

In [44]:b = np.arange(10, 19).reshape(3, 3)
b

Out[44]:array([[10, 11, 12],
       [13, 14, 15],
       [16, 17, 18]])

In [45]:c = np.arange(3)
c

Out[45]:array([0, 1, 2])

In [46]:np.dot(a, b) #　aとbの行列積

Out[46]:array([[ 45,  48,  51],
       [162, 174, 186],
       [279, 300, 321]])

In [47]:np.dot(a, c) # aとcの行列積、cは列ベクトルとみなされる

Out[47]:array([ 5, 14, 23])

In [48]:np.dot(c, a) # cとaの行列積、cは行ベクトルとみなされる

Out[48]:array([15, 18, 21])


In [49]:a = np.arange(9).reshape(3, 3)
b = np.arange(10, 19).reshape(3, 3) # aとbは前のセッションと同じ設定
a * b # 要素ごとの積

Out[49]:array([[  0,  11,  24],
       [ 39,  56,  75],
       [ 96, 119, 144]])