Advanced NumPy

Operations on NumPy arrays

Learning Objectives

Learner will explain the difference between element-wise and matrix product of two arrays.
Learner will apply reduction functions (mean, min, max) along a given axis.
Learner will be able to find a specialised numerical algorithm from the ones available in numpy.
Learner will be able to sort array along given axis.

Multiplication of two arrays is elementwise. For example, to calculate a square of each element we may use:

>>> a = np.arange(3)
>>> a
array([0, 1, 2])
>>> b = a * a
>>> b
array([0, 1, 4])

Matrix products are calculated using the np.dot function:

>>> np.dot(a, a)
5

For 1-D arrays the same result can be obtained by:

>>> np.sum(a * a)
5

Axis-based reductions

The np.sum function sums all elements regardless of the number of array dimensions:

>>> b = np.arange(9).reshape(3,3)
>>> b
array([[0, 1, 2],
       [3, 4, 5],
       [6, 7, 8]])
>>> np.sum(b)
36

If you want to sum only columns or rows, you need to pass the index of the axis over which you want to sum:

>>> np.sum(b, 0)
array([ 9, 12, 15])
>>> np.sum(b, 1)
array([ 3, 12, 21])

Other similar reduction functions are np.min, np.max or np.mean:

>>> np.min(b)
0
>>> np.min(b, 0)
array([0, 1, 2])
>>> np.min(b, 1)
array([0, 3, 6])

You can also find the index of the minimum element in each axis:

>>> np.argmin(b, 0)
array([0, 0, 0])

Sorting

NumPy also implement various sorting algorithms. To sort elements of an array you can use np.sort functions:

>>> a = np.random.rand(4)
>>> a
array([ 0.9490829 ,  0.07528673,  0.17463988,  0.95964801])
>>> np.sort(a)
array([ 0.07528673,  0.17463988,  0.9490829 ,  0.95964801])

Similarly to the reduction functions, you can also pass the axis index to sort along:

>>> b = a.reshape(2, 2)
>>> b
array([[ 0.9490829 ,  0.07528673],
       [ 0.17463988,  0.95964801]])
>>> np.sort(b, 0)
array([[ 0.17463988,  0.07528673],
       [ 0.9490829 ,  0.95964801]])
>>> np.sort(b, 1)
array([[ 0.07528673,  0.9490829 ],
       [ 0.17463988,  0.95964801]])

np.argsort returns the order of elements in a sorted array:

>>> np.argsort(a)
array([1, 2, 0, 3])

Special modules

NumPy also provides extra modules implementing basic numerical methods:

np.linalg – linear algebra,
np.fft – fast Fourier transform,
np.random – random number generators.

Finding closest element

Generate a 10 x 3 array of random numbers (using np.random.rand). From each row, find the column index of the element closest to 0.75. Make use of np.abs and np.argmin. The result should be a one-dimensional array of integers from 0 to 2.

Solving linear equations

Solve the following system of linear equations using np.linalg.solve. Test the solution. \[2x + 3y = 3\] \[5x - y = 6\]