[Python-Numpy] #3

import numpy as np

유니버설함수

arr = np.arange(10)
arr
# <출력> array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])

 

 

• 단항 유니버설 함수

np.sqrt(arr)
# <출력> 
# array([0.        , 1.        , 1.41421356, 1.73205081, 2.        ,
#        2.23606798, 2.44948974, 2.64575131, 2.82842712, 3.        ])

np.exp(arr)
# <출력> 
# array([1.00000000e+00, 2.71828183e+00, 7.38905610e+00, 2.00855369e+01,
#        5.45981500e+01, 1.48413159e+02, 4.03428793e+02, 1.09663316e+03,
#        2.98095799e+03, 8.10308393e+03])

np.exp(1)
# <출력> 2.718281828459045

 

• 이항 유니버설 함수

x = np.random.randn(8)
y = np.random.randn(8)

x 
# <출력>
# array([-0.92636774, -1.03234076,  0.88873568, -2.62295015, 
#       -0.06574232, -0.73575235,  1.52419632, -0.54659848]) 

y 
# <출력>
# array([-0.86323147,  1.42525731,  0.95813418,  1.06844799,
#       -0.80584912,  0.87617239,  0.69338792, -0.33904493])

np.maximum(x, y)
# <출력> 
# array([-0.86323147,  1.42525731,  0.95813418,  1.06844799,
#       -0.06574232,  0.87617239,  1.52419632, -0.33904493])

 

 

▷

 여러개의 배열을 반환하는 유니버설 함수
 modf 는 파이썬 내장 함수인 divmod 의 벡터화 버전. 분수를 받아서 몫과 나머지를 함께 반환

arr = np.random.randn(7) * 5
arr
# <출력> 
# array([-4.61559807, -6.27729483, -2.73862412, -1.6747488 ,
#         1.84579876, -3.88012698, -6.65983194])

remainder, whole_part = np.modf(arr)
remainder
# <출력>
# array([-0.61559807, -0.27729483, -0.73862412, -0.6747488 ,
#         0.84579876, -0.88012698, -0.65983194])

whole_part
# <출력> array([-4., -6., -2., -1.,  1., -3., -6.])

• remainde
• whole_part.

 

▷

선택적으로 out 인자 사용하여 계산 결과를 별도 저장

arr
# <출력> 
# array([-3.06515156, -2.37130948,  1.29910599,  8.06321282,
#         5.04210424,  2.07303572,  1.19492223])

np.sqrt(arr)
# <출력> 
# <input>:1: RuntimeWarning: invalid value encountered in sqrt
# array([       nan,        nan, 1.13978331, 2.83957969,
#        2.24546303,  1.43980405, 1.09312498])

np.sqrt(arr, arr)   #SQRT arr결과값을 arr에 저장
# <input>:1: RuntimeWarning: invalid value encountered in sqrt
# array([       nan,        nan, 1.13978331, 2.83957969,
#        2.24546303,  1.43980405, 1.09312498])

<input>:1: RuntimeWarning: invalid value encountered in sqrt

: 음수값은 sqrt 계산이 안되서 warning이 뜨지만 nan값으로 반환된다

 

 

 

 

▷

값이 놓여 있는 그리드에서 sqrt(x^2 + y^2)를 계산

points = np.arange(-5, 5, 0.01)           # 1000 equally spaced points
xs, ys = np.meshgrid(points, points)
ys
xs

np.meshgrid 함수는 두개의 1 차원 배열을 받아서 가능한 모든 (x, y)짝을 만들 수 있는 2 차원 배열 두개를 반환

 

 

 

▷

그리드 상의 두 포인트로 간단하게 계산 적용 가능

z = np.sqrt(xs ** 2 + ys ** 2)
z

 

matplotlib  : 파이썬을 이용한 데이터 시각화

import matplotlib.pyplot as plt

▼

plt.imshow(z, cmap=plt.cm.gray); plt.colorbar()

▼

 

▼

 

 

 

plt.title("Image plot of $\sqrt{x^2 + y^2}$ for a grid of values")

▼

 

 

 

 

plt.draw()

re-draw. interactive mode 에서 data 나 format 을 바꾸어야 할 때 사용

 

 

plt.close('all')

 

 

 

 

▷

np.where

• cond 값이 True 일때 xarr / False > yarr

xarr = np.array ([1.1,1.2,1.3,1.4,1.5])
yarr = np.array([2.1,2.2,2.3,2.4,2.5])
cond = np.array ([True,False,True,True,False])

result = [(x if c else y) for x, y, c in zip(xarr,yarr,cond)]
result
# <출력> [1.1, 2.2, 1.3, 1.4, 2.5]

 

 

• 활용

result = np.where(cond,xarr,yarr)
result 
# <출력> [1.1, 2.2, 1.3, 1.4, 2.5]

 

 

• where 은 다른 배열에 기반한 새로운 배열을 생성

arr = np.random.randn(4, 4)

arr
# <출력> 
# array([[ 0.20343263, -1.24549135,  0.04353215,  1.76619998],
#        [ 0.24949921,  1.54354193, -0.39978553,  0.02376734],
#        [ 1.38698041, -0.32689699,  0.18724001,  0.41467767],
#        [ 0.40090677, -0.19479791,  0.40132263, -0.34850082]])

arr > 0
# <출력> 
# array([[ True, False,  True,  True],
#        [ True,  True, False,  True],
#        [ True, False,  True,  True],
#        [ True, False,  True, False]])

np.where(arr > 0, 2, -2)
# <출력> array([[ 2, -2,  2,  2],
#        [ 2,  2, -2,  2],
#        [ 2, -2,  2,  2],
#        [ 2, -2,  2, -2]])

 

 

• arr 의 모든 양수를 2 로 바꿈

arr
# <출력>
# array([[ 0.20343263, -1.24549135,  0.04353215,  1.76619998],
#        [ 0.24949921,  1.54354193, -0.39978553,  0.02376734],
#        [ 1.38698041, -0.32689699,  0.18724001,  0.41467767],
#        [ 0.40090677, -0.19479791,  0.40132263, -0.34850082]])

np.where(arr > 0, 2, arr)
# <출력>
# array([[ 2.        , -1.24549135,  2.        ,  2.        ],
#        [ 2.        ,  2.        , -0.39978553,  2.        ],
#        [ 2.        , -0.32689699,  2.        ,  2.        ],
#        [ 2.        , -0.19479791,  2.        , -0.34850082]])

arr
#<출력>
# array([[ 0.20343263, -1.24549135,  0.04353215,  1.76619998],
#       [ 0.24949921,  1.54354193, -0.39978553,  0.02376734],
#       [ 1.38698041, -0.32689699,  0.18724001,  0.41467767],
#       [ 0.40090677, -0.19479791,  0.40132263, -0.34850082]])

np.where => 원본이 바뀌진않음  처음 불러왔던arr = np.random.randn(4, 4) 가 결과값으로 나온다.