6.
NUMPY − ARRAY CREATION ROUTINES NumPy
A new ndarray object can be constructed by any of the following array creation routines or
using a low-level ndarray constructor.
[Link]
It creates an uninitialized array of specified shape and dtype. It uses the following constructor:
[Link](shape, dtype=float, order='C')
The constructor takes the following parameters.
Shape Shape of an empty array in int or tuple of int
Dtype Desired output data type. Optional
Order 'C' for C-style row-major array, 'F' for FORTRAN style column-major array
Example 1
The following code shows an example of an empty array.
import numpy as np
x = [Link]([3,2], dtype=int)
print x
The output is as follows:
[[22649312 1701344351]
[1818321759 1885959276]
[16779776 156368896]]
Note: The elements in an array show random values as they are not initialized.
22
NumPy
[Link]
Returns a new array of specified size, filled with zeros.
[Link](shape, dtype=float, order='C')
The constructor takes the following parameters.
Shape Shape of an empty array in int or sequence of int
Dtype Desired output data type. Optional
Order 'C' for C-style row-major array, 'F' for FORTRAN style column-major array
Example 2
# array of five zeros. Default dtype is float
import numpy as np
x = [Link](5)
print x
The output is as follows:
[ 0. 0. 0. 0. 0.]
Example 3
import numpy as np
x = [Link]((5,), dtype=[Link])
print x
Now, the output would be as follows:
[0 0 0 0 0]
23
NumPy
Example 4
# custom type
import numpy as np
x = [Link]((2,2), dtype=[('x', 'i4'), ('y', 'i4')])
print x
It should produce the following output:
[[(0, 0) (0, 0)]
[(0, 0) (0, 0)]]
[Link]
Returns a new array of specified size and type, filled with ones.
[Link](shape, dtype=None, order='C')
The constructor takes the following parameters.
Shape Shape of an empty array in int or tuple of int
Dtype Desired output data type. Optional
Order 'C' for C-style row-major array, 'F' for FORTRAN style column-major array
Example 5
# array of five ones. Default dtype is float
import numpy as np
x = [Link](5)
print x
The output is as follows:
24
NumPy
[ 1. 1. 1. 1. 1.]
Example 6
import numpy as np
x = [Link]([2,2], dtype=int)
print x
Now, the output would be as follows:
[[1 1]
[1 1]]
25
7. NUMPY − ARRAY FROM EXISTING DATA NumPy
In this chapter, we will discuss how to create an array from existing data.
[Link]
This function is similar to [Link] except for the fact that it has fewer parameters. This
routine is useful for converting Python sequence into ndarray.
[Link](a, dtype=None, order=None)
The constructor takes the following parameters.
Input data in any form such as list, list of tuples, tuples, tuple of tuples or tuple
a
of lists
dtype By default, the data type of input data is applied to the resultant ndarray
order C (row major) or F (column major). C is default
The following examples show how you can use the asarray function.
Example 1
# convert list to ndarray
import numpy as np
x = [1,2,3]
a = [Link](x)
print a
Its output would be as follows:
[1 2 3]
26
NumPy
Example 2
# dtype is set
import numpy as np
x = [1,2,3]
a = [Link](x, dtype=float)
print a
Now, the output would be as follows:
[ 1. 2. 3.]
Example 3
# ndarray from tuple
import numpy as np
x = (1,2,3)
a = [Link](x)
print a
Its output would be:
[1 2 3]
Example 4
# ndarray from list of tuples
import numpy as np
x = [(1,2,3),(4,5)]
a = [Link](x)
print a
Here, the output would be as follows:
[(1, 2, 3) (4, 5)]
27
NumPy
[Link]
This function interprets a buffer as one-dimensional array. Any object that exposes the buffer
interface is used as parameter to return an ndarray.
[Link](buffer, dtype=float, count=-1, offset=0)
The constructor takes the following parameters.
buffer Any object that exposes buffer interface
dtype Data type of returned ndarray. Defaults to float
count The number of items to read, default -1 means all data
offset The starting position to read from. Default is 0
Example 5
The following examples demonstrate the use of frombuffer function.
import numpy as np
s = 'Hello World'
a = [Link](s, dtype='S1')
print a
Here is its output:
['H' 'e' 'l' 'l' 'o' ' ' 'W' 'o' 'r' 'l' 'd']
[Link]
This function builds an ndarray object from any iterable object. A new one-dimensional array
is returned by this function.
[Link](iterable, dtype, count=-1)
Here, the constructor takes the following parameters.
iterable Any iterable object
dtype
Data type of resultant array
28