NumPy is a python library that makes it easy to perform mathematical and logical operations on arrays. It provides high performance n-dimensional array object and tools to work with the arrays.<\/p>\n\n\n\n
import numpy as np\nprint (np.__version__)\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
1.9.3\n<\/code><\/pre>\n\n\n\nCreating numpy arrays<\/h3>\n\n\n\n
There are many ways to create numpy arrays<\/p>\n\n\n\n
# create arrays\navalues = [12, 34, 45]\na = np.array(avalues)\nb = np.array([11.1,22.2,33.3])\n\n# 2D array, matrix\nc = np.array([(1, 2, 3), (4 , 5, 6)])\n\n# print results\nprint('a = ', a) \nprint('b = ', b) \nprint(c)\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
a = [12 34 45]\nb = [ 11.1 22.2 33.3]\n[[1 2 3]\n [4 5 6]] \n<\/code><\/pre>\n\n\n\nUsing arange() function to create arrays<\/p>\n\n\n\n
d = np.arange(5) # create array of 5 numbers starting from 0\ne = np.arange(9)+1 # create array of 9 numbers starting from 1\nf = np.arange(10, 20) # array of numbers 10 to 19\ng = np.arange(10, 31, 5) # array 10 to 30 with 5 increments\nh = len(np.arange(10,15)) # get size of array\ni = np.arange(10, 20).size # get size of array\nj = np.arange(9).reshape(3,3) # create 2D array\n\n# print results\nprint('d = ', d)\nprint('e = ', e)\nprint('f = ', f)\nprint('g = ', g)\nprint('h = ', h)\nprint('i = ', i)\nprint(j)\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
d = [0 1 2 3 4]\ne = [1 2 3 4 5 6 7 8 9]\nf = [10 11 12 13 14 15 16 17 18 19]\ng = [10 15 20 25 30]\nh = 5\ni = 10\n[[0 1 2]\n [3 4 5]\n [6 7 8]]\n<\/code><\/pre>\n\n\n\nUsing numpy function linspace()<\/p>\n\n\n\n
d = np.linspace(10, 20, 5) # create elements with 2.5 intervals\ne = np.linspace(10, 20, 5, retstep=True) # also shows the interval size\nprint(d)\nprint(e)\nprint(e[1]) # get the interval size\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
[ 10. 12.5 15. 17.5 20. ]\n(array([ 10. , 12.5, 15. , 17.5, 20. ]), 2.5)\n2.5\n<\/code><\/pre>\n\n\n\nUsing zeros() and ones() functions<\/p>\n\n\n\n
a = np.zeros(4) # array of 4 zero elements\nb = np.ones(4) # array of 4 one elements\nc = np.ones(5, dtype='int64') # create int64 array\nd = np.zeros((4,3)) # 4 x 3 matrix of zeros\ne = np.zeros((2,3,4)) # Two 3 x 4 matrices of zeros\nprint(a, \"\\n\", b, \"\\n\", c, \"\\n-----\\n\", d, \"\\n-----\\n\", e)\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
[ 0. 0. 0. 0.] \n [ 1. 1. 1. 1.] \n [1 1 1 1 1] \n-----\n[[ 0. 0. 0.]\n [ 0. 0. 0.]\n [ 0. 0. 0.]\n [ 0. 0. 0.]] \n-----\n [[[ 0. 0. 0. 0.]\n [ 0. 0. 0. 0.]\n [ 0. 0. 0. 0.]]\n\n [[ 0. 0. 0. 0.]\n [ 0. 0. 0. 0.]\n [ 0. 0. 0. 0.]]]\n<\/code><\/pre>\n\n\n\nArray Type<\/h3>\n\n\n\n
A numpy array can only have one type. To find the type of an array,<\/p>\n\n\n\n
print(a.dtype) # int64\nprint(b.dtype) # float64\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
type of a = int64\ntype of b = float64\n<\/code><\/pre>\n\n\n\nSince every element of a numpy array must have the same data type, in the following example, int and float elements will be converted to complex numbers:<\/p>\n\n\n\n
tpl = (10, -1.23, 2+3j)\nc = np.array(tpl)\nprint(c)\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
[ 10.00+0.j -1.23+0.j 2.00+3.j]\n<\/code><\/pre>\n\n\n\nUsing arrays<\/h3>\n\n\n\n
Working with 1D array<\/p>\n\n\n\n
a = np.array([-3, -2, 0, 1, 2])\nprint(a)\nprint(a[1])\na[2] = 100\nprint(a)\nprint(a.size)\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
[-3 -2 0 1 2]\n-2\n[ -3 -2 100 1 2]\n5\n<\/code><\/pre>\n\n\n\nWorking with 2D array<\/p>\n\n\n\n
b = np.arange(21)\nb.shape = (3, 7)\nprint(b)\nprint('-----')\nprint(b[2])\nprint(b[2,1])\nprint(b[2][1])\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
[[ 0 1 2 3 4 5 6]\n [ 7 8 9 10 11 12 13]\n [14 15 16 17 18 19 20]]\n-----\n[14 15 16 17 18 19 20]\n15\n15\n<\/code><\/pre>\n\n\n\nWorking with 3D array<\/p>\n\n\n\n
c = np.arange(27)\nc.shape = (3,3,3)\nprint(c)\nprint('-----')\nprint(c[0])\nprint('-----')\nprint(c[0,1])\nprint('-----')\nprint(c[0,1,2])\nprint('-----')\nc[0,1,2] = 999\nprint(c)\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
[[[ 0 1 2]\n [ 3 4 5]\n [ 6 7 8]]\n\n [[ 9 10 11]\n [12 13 14]\n [15 16 17]]\n\n [[18 19 20]\n [21 22 23]\n [24 25 26]]]\n-----\n[[0 1 2]\n [3 4 5]\n [6 7 8]]\n-----\n[3 4 5]\n-----\n5\n-----\n[[[ 0 1 2]\n [ 3 4 999]\n [ 6 7 8]]\n\n [[ 9 10 11]\n [ 12 13 14]\n [ 15 16 17]]\n\n [[ 18 19 20]\n [ 21 22 23]\n [ 24 25 26]]]\n<\/code><\/pre>\n\n\n\nArithmetic operations on arrays<\/h3>\n\n\n\n
Note that numpy overloads operators operate differently on python arrays and numpy arrays.<\/p>\n\n\n\n
# subtract all values\nprint(a - b) # [ 0.9 11.8 11.7]\n# cube all values\nprint(a**3) # [ 1728 39304 91125]\nprint(2 * np.cos(a)) # [ 1367.631 10941.048 36926.037]\nprint(a < 30) # [ True False False]\n# multiply values\nprint(a * b) # [ 133.2 754.8 1498.5]\n# matrix product\nprint(a.dot(b)) # 2386.5\n<\/code><\/pre>\n\n\n\nOutput is in the code comments.<\/p>\n\n\n\n
Special Functions<\/h3>\n\n\n\nimport numpy as np\nimport matplotlib as pyplot as plt\nx = np.arange(0,3*np.pi, 0.1)\ny = np.sin(x)\nplt.plot(x,y)\nplt.show\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
You should see a sign function\n<\/code><\/pre>\n\n\n\nYou can also use cos() and tan() in a similar way. The following function calculates the e value.<\/p>\n\n\n\n
import numpy as np\nx = np.array([1,2,3])\nprint(np.exp(x))\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
[ 2.71828183 7.3890561 20.08553692 ]\n<\/code><\/pre>\n\n\n\nThe following code calculates the natural log:<\/p>\n\n\n\n
import numpy as np\ny = np.array([1,2,3])\nprint(np.log(ar))\n<\/code><\/pre>\n\n\n\noutput<\/p>\n\n\n\n
[ 0. 0.693141718 1.09861229 ]\n<\/code><\/pre>\n\n\n\nSimilarly log10 with give log base 10 and so on. numpy packs more power than what we covered here.<\/p>\n","protected":false},"excerpt":{"rendered":"
NumPy is a python library that makes it easy to perform mathematical and logical operations on arrays. It provides high performance n-dimensional array object and tools to work with the arrays. Why use numpy instead of a python list Takes less memory Faster than python lists More powerful Easier to use What is my numpy […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[203],"tags":[208,137],"class_list":["post-721","post","type-post","status-publish","format-standard","hentry","category-python","tag-numpy","tag-python"],"_links":{"self":[{"href":"https:\/\/molecularsciences.org\/content\/wp-json\/wp\/v2\/posts\/721","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/molecularsciences.org\/content\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/molecularsciences.org\/content\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/molecularsciences.org\/content\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/molecularsciences.org\/content\/wp-json\/wp\/v2\/comments?post=721"}],"version-history":[{"count":1,"href":"https:\/\/molecularsciences.org\/content\/wp-json\/wp\/v2\/posts\/721\/revisions"}],"predecessor-version":[{"id":722,"href":"https:\/\/molecularsciences.org\/content\/wp-json\/wp\/v2\/posts\/721\/revisions\/722"}],"wp:attachment":[{"href":"https:\/\/molecularsciences.org\/content\/wp-json\/wp\/v2\/media?parent=721"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/molecularsciences.org\/content\/wp-json\/wp\/v2\/categories?post=721"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/molecularsciences.org\/content\/wp-json\/wp\/v2\/tags?post=721"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}