# coding: utf-8

# # Matrix norms

# In[1]:

import numpy as np
import numpy.linalg as la
import matplotlib.pyplot as pt


# Here's a matrix of which we're trying to compute the norm:

# In[14]:

n = 2
A = np.random.randn(n, n)


# Recall:
# 
# $$||A||=\max_{\|x\|=1} \|Ax\|,$$
# 
# where the vector norm must be specified, and the value of the matrix norm $\|A\|$ depends on the choice of vector norm.
# 
# For instance, for the $p$-norms, we often write:
# 
# $$||A||_2=\max_{\|x\|=1} \|Ax\|_2,$$
# 
# and similarly for different values of $p$.

# --------------------
# We can approximate this by just producing very many random vectors and evaluating the formula:

# In[16]:

xs = np.random.randn(n, 1000)


# First, we need to bring all those vectors to have norm 1. First, compute the norms:

# In[32]:

p = 2
norm_xs = np.sum(np.abs(xs)**p, axis=0)**(1/p)
norm_xs.shape


# Then, divide by the norms and assign to `normalized_xs`:

# In[33]:

normalized_xs = xs/norm_xs
la.norm(normalized_xs[:, 316], p)


# Let's take a look:

# In[34]:

pt.plot(normalized_xs[0], normalized_xs[1], "o")
pt.gca().set_aspect("equal")


# Now apply $A$ to these normalized vectors:

# In[35]:

A_nxs = A.dot(normalized_xs)


# --------------
# Let's take a look again:

# In[36]:

pt.plot(normalized_xs[0], normalized_xs[1], "o", label="x")
pt.plot(A_nxs[0], A_nxs[1], "o", label="Ax")
pt.legend()
pt.gca().set_aspect("equal")


# Next, compute norms of the $Ax$ vectors:

# In[ ]:

norm_Axs = np.sum(np.abs(A_nxs)**p, axis=0)**(1/p)
norm_Axs.shape


# What's the biggest one?

# In[23]:

np.max(norm_Axs)


# Compare that with what `numpy` thinks the matrix norm is:

# In[24]:

la.norm(A, p)


# In[9]:

A = np.arange(9).reshape(3,3)
A


# In[13]:

np.sum(A)


# In[ ]: