统计学习方法 | 工具包及常用函数笔记 [持续更新]

科学计算

NumPy

• pi = π
• sin()
• poly1d(c_or_r, r=False, variable=None)
  • poly1d([1,2,3]) = $ 1x^2 + 2x^1 + 3x^0 $
• linspace(start, stop, num=50, endpoint=True, retstep=False, dtype=None, axis=0) 生成均匀分布的点
• random.normal(loc=0.0, scale=1.0, size=None) 生成正态分布随机数
• append(arr, values, axis=None) 将values附加到array的最后
  • axis为None时返回一维数组；为0时要求列数相同；为1时要求行数相同

import numpy as np
# pi
c = np.pi
print('pi: ', c)
# sin()
s = np.sin(np.pi / 2)
print('sin: ', s)

pi:  3.141592653589793
sin:  1.0

# poly1d()
p1 = np.poly1d([1, 2, 3])
print(p1)
print('x=0.5: ', p1(0.5))
p2 = np.poly1d([1, 2, 3], True)
print(p2)
print('x=0.5: ', p2(0.5))

   2
1 x + 2 x + 3
x=0.5:  4.25
   3     2
1 x - 6 x + 11 x - 6
x=0.5:  -1.875

# linspace()
x_points = np.linspace(2.0, 4.0, num=5)
print('x points: ', x_points)

# random.normal()
mu, sigma = 0, 0.1 # mean and standard deviation
s = np.random.normal(mu, sigma, 10)
print('normal distribution randoms: ', s)

x points:  [2.  2.5 3.  3.5 4. ]
normal distribution randoms:  [ 0.0970498   0.10999027 -0.07687955  0.05997433 -0.06545673 -0.03728823
 -0.10978893  0.03702012 -0.09140556  0.15170791]

# append()
# axis=None
print('axis=None: ')
print(np.append([1, 2, 3], [[4, 5, 6], [7, 8, 9]]))
# axis=0
print('axis=0: ')
print(np.append([[1, 2, 3], [4, 5, 6]], [[7, 8, 9]], axis=0))
# axis=1
print('axis=1: ')
print(np.append([[1,2,3], [4,5,6]], \
                           [[7,8,9], [10,11,12]], axis=1))

axis=None: 
[1 2 3 4 5 6 7 8 9]
axis=0: 
[[1 2 3]
 [4 5 6]
 [7 8 9]]
axis=1: 
[[ 1  2  3  7  8  9]
 [ 4  5  6 10 11 12]]

SciPy (Based on NumPy)

- optimize

• optimize.leastsq(func, x_0, args=())

可视化

matplotlib

- pyplot

• pyplot.plot(x_points, y_points, label)
• pyplot.legend() 添加图例