Numpyfunctioninpythonforadvanceddata.pptx

LEARNING OBJECTIVES
•
•
•
•
•
•
•
•
•
Introduction to
NumPy NumPy Array
Creating NumPy
Array Array
Attributes
Array Methods
Array Indexing
Slicing Arrays
Array
Operation
Iteration
through Arrays
At the end of this session, you will learn:

INTRODUCTION TO NUMPY
NumPy stands for ‘Numeric Python’
•
4
'Numerical Python’
Used for mathematical and scientific computations
• Also provides ‘linalg’ module which contains functions like det, eig,
norm to apply linear algebra on NumPy arrays
• NumPy array is the most widely used object of the NumPy
library

INTRODUCTION TO NUMPY
• Installing NumPy
Use the following command to install Numpy using jupyter
notebook
• Importing numpy as alias np is a common
practice
5

NUMPY ARRAY
Looks similar to a
list
It is a grid of values, indexed by positive
integers
It generally contains numeric values. However it can also contain
strings
Works faster than lists because it is
homogeneous
It can be
multidimensional
7

1D NUMPY ARRAY
• One dimensional array contains elements only in one dimension. In other
words, the shape of the numpy array should contain only one value in the
tuple
8

2D NUMPY ARRAY
• Two dimensional array is an array within an
array
• The position of an data element is referred by two indices instead
of one
9

3D NUMPY ARRAY
• A three-dimensional (3D) array is composed of 3 nested levels of arrays,
one for each dimension
10

CONVERTING A LIST INTO NUMPY ARRAY
• np.array() is used to create a numpy array from a
list
12

CREATING NUMPY ARRAY
• Numpy arrays be used to create array of strings as
well
13

NUMPY ARRAY OF RANDOM NUMBERS
• Create an array of 20 random numbers using random() method
from the random module
The required number of random numbers
is passed through the ‘size’ parameter
random() method returns random numbers
over the half-open interval [0.0, 1.0)
14

• rand() method creates an array of random numbers of the given
shape and between (0, 1)
• The dimensions of the returned array, should all be
positive
• If no argument is given a single Python float is
returned
15

• The randn() returns a set of values from the standard normal
distribution
• The dimensions of the returned array, should all be
positive
• If no argument is given a single Python float is
returned
16

• The randint() returns random integers from low (inclusive) to high
(exclusive)
17

CREATING NUMPY ARRAY USING ARANGE()
• np.arange() can also be used to create a NumPy
array
• The numbers generated have the same
difference
• The function generates as many possible numbers in the given
range
Numpy.arange (start, stop, step, dtype)
The function name
The start of the
interval (optional).
Default is 0
The end of the
interval
The “step” between
values (optional)
The data type (optional)
18

CREATING NUMPY ARRAY USING ARANGE()
• The np.arange() create a series of values from 10 to 100
with a difference of 2, stored as a numpy array
19

CREATING NUMPY ARRAY USING LINSPACE()
• linspace() generates a specified number of values in a specified
range
Syntax:
numpy.linspace ( start, stop , num, dtype )
The start of the
interval
(optional)
The end of
the interval
The number
of values
required
in the
interval
The data
type
(optional
)
20

CREATING NUMPY ARRAY USING LINSPACE()
• np.linspace() produces a sequence of 10 evenly spaced values from
1 to 100, stored as a numpy array
21

CREATING NUMPY ARRAY OF ZEROES
• Creating 1D numpy array of
zeros
• Creating 1D numpy array of
ones
22

CREATING 2D NUMPY ARRAY
• np.empty() returns the matrix with arbitrary values of given
shape and data type
• ‘dtype = object’ returns None
values
23

• np.full() returns the matrix of given shape with the value set
by the ‘fill_value’ parameter
24

• np.identity() returns the identity matrix of specified
shape
25

• np.eye() creates NxM matrix with value ‘1’ on the k-th diagonal
and remaining entries as zero
•
•
K > 0 represents upper
diagonal K < 0 represents
lower diagonal
K = 0 represents main
diagonal
K > 0 represents upper
diagonal
K < 0 represents lower
diagonal
26

ARRAY ATTRIBUTES
Some of the attributes of the numpy
array are:
shape size dtype
ndim
Attributes are the features/characteristics of an object that describes the object
Attributes do not have parentheses following them
28

ARRAY ATTRIBUTES-SHAPE
• The shape returns the number of rows and columns of the array
respectively
29

ARRAY ATTRIBUTES-SIZE
• The size returns the number of elements in an
array
30

ARRAY ATTRIBUTES-DTYPE
• The dtype returns the type of the data along with the size in
bytes
In this example, the array consists of 64-bit floating-point numbers. Thus, the dtype of the array is
float64
31

ARRAY ATTRIBUTES-NDIM
• The ndim returns the number of axes (dimension) of the
array
32

ARRAY METHODS
Methods are object functions that takes
parameters in the parentheses and returns
the modified object
35

ARRAY METHODS
• In this example, the reshape(6, 1) is reshaping a 3 X 2 array to 6 X 1
array
36

INDEXING ARRAY
• The element in the array can be accessed by the positional index of the
element
• The index for an array starts at 0 from left and at -1 starts from
the right
39

SLICING 1D ARRAY
• Slicing allows us to access more than one
element
41

SLICING 2D ARRAY
• Slicing in 2d array returns a sub-matrix of the original
matrix
42

The index returns an element of the array, the slice returns a list of elements.
43

If you try to add arrays with different dimensions, you get an error.
44

ARITHMETIC OPERATIONS
Addition and Subtraction of 1D
array
46

Multiply each element in the array
by 2
47

Element-wise multiplication of two 3x3
matrices
m
1
m
2
m1
*
m2
This product is also known as ‘Hadamard Product’
48

Matrix multiplication of two 3x3
matrices
49

• The min() returns the minimum value present in the
array
• The max() returns the maximum value present in the
array
50

• The var() returns the variance of all the elements in the
array
• The std() returns the standard deviation of all the elements in the
array
51

• The np.square() returns the square of the
elements
• The np.power() is used to raise the numbers in the array to the given
value
52

• The np.transpose() reverses the dimension of the
array
53

CONCATENATE 1D ARRAY
• Two or more arrays will get joined along existing (first) axis, provided
they have the same shape
54

CONCATENATE 2D ARRAY
• We can concatenate 2D arrays either along rows (axis = 0) or columns
(axis
= 1), provided they have same shape
55

We can not concatenate the arrays with different dimensions
56

FLATTEN THE ARRAY
• The flatten() function collapses the original array into a single
dimension
57

Reshape the Array
•
•
•
Reshaping means changing the shape of an array.
The shape of an array is the number of elements in each dimension.
By reshaping we can add or remove dimensions or change the number of
elements in each dimension.
1 2 3 4 5 6 7 8 9 10 11 12
1 2 3
4 5 6
7 8 9
10 11 12
58

Vertical Stack (vstack)
•
•
•
You can do vertical staking for one vector (vstack)
If you want to perform for more than one vector you want to mention in the
list
If you are vertically stacking more than one array the size of array should be
same
59

Horizontal Stack (hstack)
•
•
•
You can do horizontal staking for one vector (hstack)
If you want to perform for more than one vector you want to mention in the list
If you are horizontally stacking more than one array the size of array may be
different not an issue
60

ITERATING THROUGH 1D ARRAY
• The for loop can be used to iterate through the array
elements
62

ITERATING THROUGH 2D ARRAY USING NESTED FOR LOOP
• To print each element in the 2D array, use nested for
loop
64

NumPy arrays are faster than list. The below code shows that NumPy arrays are very much
faster than the lists in python.
65

Numpyfunctioninpythonforadvanceddata.pptx

More Related Content

Similar to Numpyfunctioninpythonforadvanceddata.pptx

Recently uploaded

Numpyfunctioninpythonforadvanceddata.pptx