Scientific Computing with Python - NumPy | WeiYuan

Scientific Computing with Python - NumPy 2017/08/03 (Thus.) WeiYuan

site: v123582.github.io line: weiwei63 § 全端⼯程師 + 資料科學家略懂⼀點網站前後端開發技術，學過資料探勘與機器學習的⽪⽑。平時熱愛參與技術社群聚會及貢獻開源程式的樂趣。

Outline § About NumPy § Ndarray § Create a new Array § Property of Array § Operation of Array § Matrix § Advanced Usages 3

the Ecosystem of Python 5Reference: https://www.edureka.co/blog/why-you-should-choose-python-for-big-data

6Reference: https://www.slideshare.net/gabrielspmoreira/python-for-data-science-python-brasil-11-2015

About NumPy § NumPy is the fundamental package for scientific computing with Python. It contains among other things: • a powerful N-dimensional array object • sophisticated (broadcasting) functions • tools for integrating C/C++ and Fortran code • useful linear algebra, Fourier transform, and random number capabilities • be used as an efficient multi-dimensional container of generic data. 7

About NumPy § NumPy is the fundamental package for scientific computing with Python. It contains among other things: • a powerful N-dimensional array object • sophisticated (broadcasting) functions • tools for integrating C/C++ and Fortran code • useful linear algebra, Fourier transform, and random number capabilities • be used as an efficient multi-dimensional container of generic data. 8

Try it! § #練習：Import the numpy package under the name np 9

Try it! § #練習：Print the numpy version and the configuration 10

Ndarray § shape § ndim § dtype § size § itemsize § data 12 1 2 3 4 5 6 7 8 9 10 11 12 from numpy import * a = array([[ 0, 1, 2, 3, 4], [ 5, 6, 7, 8, 9], [10, 11, 12, 13, 14] ])

Ndarray § shape § ndim § dtype § size § itemsize § data 13 1 2 3 4 5 6 7 8 9 10 11 12 from numpy import * a = array([[ 0, 1, 2, 3, 4], [ 5, 6, 7, 8, 9], [10, 11, 12, 13, 14] ]) a.shape # (3, 5) a.ndim # 2 a.dtype.name # 'int32’ a.size # 15 a.itemsize # 4 type(a) # numpy.ndarray

Create a new Ndarray § One Dimension Array (1dnarray) § Multiple Dimension Array (ndarray) § Zeros, Ones, Empty § arange and linspace § random array § array from list/tuple 15

Create a new Ndarray § One Dimension Array (1dnarray) 16 1 2 3 4 5 6 7 8 9 import numpy as np arr1 = np.array([0, 1, 2, 3, 4]) # array([0 1 2 3 4]) type(arr1) # <type 'numpy.ndarray'> arr1.dtype # dtype('int64') array( )

Create a new Ndarray § One Dimension Array (1dnarray) 17 1 2 3 4 5 6 7 8 9 import numpy as np arr1 = np.array([0, 1, 2, 3, 4]) # array([0 1 2 3 4]) type(arr1) # <type 'numpy.ndarray'> arr1.dtype # dtype('int64') arr2 = np.array([1.2, 2.4, 3.6]) # array([1.2, 2.4, 3.6]) type(arr2) # <type 'numpy.ndarray'> arr2.dtype # dtype('float64') array( )

Create a new Ndarray § Question：How to assign data type for an array ? 18

Create a new Ndarray § Multiple Dimension Array (ndarray) 19 1 2 3 4 5 6 7 8 9 import numpy as np a = np.array([[1, 2, 3], [4, 5, 6]]) # array([[1, 2, 3], # [4, 5, 6]]) array( )

Create a new Ndarray § Question：How to change shape from 1-d array ? 20

Create a new Ndarray § Zeros, Ones, Empty 21 1 2 3 4 5 6 7 8 9 import numpy as np zeros = np.zeros(5) # array([ 0., 0., 0., 0., 0.]) zeros( )

Create a new Ndarray § Zeros, Ones, Empty 22 1 2 3 4 5 6 7 8 9 import numpy as np zeros = np.ones(5) # array([ 1., 1., 1., 1., 1.]) ones( )

Create a new Ndarray § Zeros, Ones, Empty 23 1 2 3 4 5 6 7 8 9 import numpy as np zeros = np.empty(5) # array([ 0., 0., 0., 0., 0.]) empty( )

Create a new Ndarray § arange and linspace 24 1 2 3 4 5 6 7 8 9 import numpy as np arange = np.arange(5) # array([0 1 2 3 4]) arange( )

Create a new Ndarray § arange and linspace 25 1 2 3 4 5 6 7 8 9 import numpy as np linspace = np.linspace(0, 4, 5) # array([ 0., 1., 2., 3., 4.]) linspace( )

Create a new Ndarray § random array 26 1 2 3 4 5 6 7 8 9 import numpy as np linspace = np.random.randint(0, 2, size=4) # array([ 0, 1, 1, 1]) random.randint( )

Try it! § #練習：Create a 3x3x3 array with random values 27

Try it! § #練習：Find indices of non-zero elements 28

Create a new Ndarray § array from list/tuple 29 1 2 3 4 5 6 7 8 9 import numpy as np x = [1,2,3] a = np.asarray(x) x = (1,2,3) a = np.asarray(x) asarray( )

Try it! § #練習：Create a null vector of size 10 30

Try it! § #練習：Create a vector with values ranging from 10 to 49 31

Try it! § #練習：Create a 3x3 identity matrix 32

Property of Ndarray § shape § ndim § dtype § size § itemsize § data 34 1 2 3 4 5 6 7 8 9 10 11 12 import numpy as np a = np.array( [[11, 12, 13, 14, 15], [16, 17, 18, 19, 20], [21, 22, 23, 24, 25], [26, 27, 28 ,29, 30], [31, 32, 33, 34, 35] ])

Property of Ndarray § shape § ndim § dtype § size § itemsize § data 35 1 2 3 4 5 6 7 8 9 10 11 12 print(type(a)) print(a.shape) print(a.ndim) print(a.dtype) print(a.size) print(a.itemsize) print(a.nbytes)

Try it! § #練習：How to find the memory size of any array 36

data type 37 § Question：How to assign data type for an array ? 1. Set dtype with create the array 2. Change dtype function

1. Set dtype with create the array 38 1 2 3 4 5 6 7 8 9 x = numpy.array([1,2.6,3], dtype = numpy.int64) print(x) # print(x.dtype) # x = numpy.array([1,2,3], dtype = numpy.float64) print(x) # print(x.dtype) # array( )

2. Change dtype function 39 1 2 3 4 5 6 7 8 9 x = numpy.array([1,2.6,3], dtype = numpy.float64) y = x.astype(numpy.int32) print(y) # [1 2 3] print(y.dtype) z = y.astype(numpy.float64) print(z) # [ 1. 2. 3.] print(z.dtype) astype( )

data shape 41 § Question：How to change shape from 1-d array ? 1. Set multiple array with create the array 2. Assign new shape to shape property 3. Change shape function

1. Set multiple array with create the array 42 1 2 3 4 5 6 7 8 9 import numpy as np a = np.array([[1,2,3],[4,5,6]]) a.shape # (2, 3) array( )

2. Assign new shape to shape property 43 1 2 3 4 5 6 7 8 9 a = np.array([[1,2,3],[4,5,6]]) a.shape = (3,2) a # [[1, 2], [3, 4], [5, 6]] array( )

3. Change shape function 44 1 2 3 4 5 6 7 8 9 a = np.array([[1,2,3],[4,5,6]]) b = a.reshape(3,2) b # [[1, 2], [3, 4], [5, 6]] reshape( )

3. Change shape function 45 1 2 3 4 5 6 7 8 9 a = np.array([[1,2,3],[4,5,6]]) b = a.reshape(3,2) b # [[1, 2], [3, 4], [5, 6]] a.resize(3,2) a # [[1, 2], [3, 4], [5, 6]] resize( )

Try it! § #練習：Create a 3x3 matrix with values ranging from 0 to 8 46

index and slicing § index § slicing 47 1 2 3 4 array[0] # 0 array[1] # 1 array[-1] # 4 1 2 3 4 5 array[1:3] # [1, 2] array[:4] # [0, 1, 3] array[3:] # [3, 4] array[1:4:2] # [1, 3] array[::-1] # [4, 3, 2, 1, 0] ([0, 1, 2, 3, 4])

index and slicing § index § slicing 48 1 2 3 4 array[1] # [0, 1] array[1][0] # 0 array[1][1] # 1 array[2][0] # 2 1 2 3 4 5 array[0:2] # [[0, 1], [2, 3]] array[:2] # [[0, 1], [2, 3]] array[2:] # [[4, 5]] ( [0, 1, 0, 1, 0], [2, 3, 2, 3, 2], [4, 5, 4, 5, 4] )

index and slicing § slicing 49 1 2 3 4 array[0, 1:4] # [1, 0, 1] array[[0, 0, 0], [1, 2, 3]] array[1:3, 0] # [1, 3, 5] array[[1, 2], [0, 0, 0]] ( [0, 1, 0, 1, 0], [2, 3, 2, 3, 2], [4, 5, 4, 5, 4] )

Try it! § #練習：Create a null vector of size 10 but the fifth value which is 1 50

Try it! § #練習：Reverse a vector (first element becomes last) 51

Try it! § #練習：Create a 2d array with 1 on the border and 0 inside 52

Try it! § #練習：Create a 8x8 matrix and fill it with a checkerboard pattern 53

Try it! § #練習： 54 1 2 3 4 5 6 7 8 9 import numpy as np a = np.array([[1,2,3],[3,4,5],[4,5,6]]) # 第二列元素 # 第二行元素 # 除了第二列的元素

Try it! § #練習： 55 1 2 3 4 array[::2,::2] array[:, 1]

sophisticated (broadcasting) 58

Advanced Usages § Boolean indexing and Fancy indexing § Boolean masking § Incomplete Indexing § Where function § Customize dtype 66

Thanks for listening. 2017/08/03 (Thus.) Scientific Computing with Python – NumPy Wei-Yuan Chang v123582@gmail.com v123582.github.io

Scientific Computing with Python - NumPy | WeiYuan

Scientific Computing with Python - NumPy | WeiYuan

More Related Content

What's hot

Similar to Scientific Computing with Python - NumPy | WeiYuan

More from Wei-Yuan Chang

Recently uploaded

In this document

Scientific Computing with Python - NumPy | WeiYuan