#!/usr/bin/env python
# coding: utf-8

# # MatRepr
# 
# Render sparse and dense matrices to HTML and Latex, with a Jupyter extension.

# In[1]:


import scipy.sparse
import numpy as np
np.random.seed(123)

# so matrepr can be imported from the source tree.
import sys
sys.path.insert(0, '..')

from matrepr import mdisplay, mprint


# ### Load MatRepr Jupyter extension

# In[2]:


get_ipython().run_line_magic('load_ext', 'matrepr')


# ## SciPy sparse matrix

# In[3]:


scipy.sparse.random(6, 6, density=0.5)


# ## 2D NumPy array

# In[4]:


mat = np.random.random((10, 10))
mdisplay(mat, floatfmt=".2f")


# In[5]:


mat = np.random.random((500, 25))
mdisplay(mat, floatfmt=".2f", max_rows=15, max_cols=25)


# ## 2D NumPy array with row and column labels

# In[6]:


cities = ["Boston", "Buffalo", "Chicago", "Cleveland", "Dallas", "Denver"]
distances = np.array([
    [None, 457, 983, 639, 1815, 1991],
    [457, None, 536, 192, 1387, 1561],
    [983, 536, None, 344, 931, 1050],
    [639, 192, 344, None, 1205, 1369],
    [1815, 1387, 931, 1205, None, 801],
    [1991, 1561, 1050, 1369, 801, None],
])

mdisplay(distances, title=None, row_labels=cities, col_labels=cities, fill_value="--")


# ## 1D arrays and vectors

# In[7]:


vec = np.random.random((1000,))
mdisplay(vec)


# In[8]:


mprint(vec, indices=False, title=False)


# ## Sparse N-dimensional Tensors
# 
# Sparse multidimensional tensors are presented as a list of tuples.

# In[9]:


import sparse
tensor_4d = sparse.random((1000, 10, 10, 10, 10), density=0.1234)
mdisplay(tensor_4d)


# In[10]:


mprint(tensor_4d)


# ## Dense N-dimensional Numpy Arrays

# In[11]:


dense_4d = np.random.random((3, 3, 3, 3))
mdisplay(dense_4d, floatfmt=".2f")


# ## Graph Adjacency Matrix with Edge and Vertex Weights

# In[12]:


import graphblas as gb

edges = gb.Matrix.from_coo(
    [3, 0, 3, 5, 6, 0, 6, 1, 6, 2, 4, 1],
    [0, 1, 2, 2, 2, 3, 3, 4, 4, 5, 5, 6],
    [3, 2, 3, 1, 5, 3, 7, 8, 3, 1, 7, 4],
    nrows=7, ncols=7
)
vertices = gb.Vector.from_coo([range(7)], [f"V{i+1}" for i in range(7)], size=7)

mdisplay(edges, row_labels=vertices, col_labels=vertices, title=False)


# ## Nested matrices

# In[13]:


# You may mix types if the datastructure allows, as a Python list does
mat = [
    [scipy.sparse.random(2, 2, density=0.8), [[1, 2], [3, 4]]],
    [np.array([[1, 2], [3, 4]]), scipy.sparse.random(2, 2, density=0.6)]
]

mdisplay(mat, floatfmt=".2f")


# ## Large matrices

# In[14]:


scipy.sparse.eye(100_000_000, format="csr")


# In[15]:


r = scipy.sparse.random(10000, 10000, density=0.23456789, format="csr")


# In[16]:


r


# ## Duplicate entries
# 
# Some sparse formats allow multiple values with the same indices.

# In[17]:


row = [0,   1,   1,   2,   2,   2]
col = [0,   1,   1,   2,   2,   2]
val = [1, 2.1, 2.2, 3.1, 3.2, 3.3]
dupes = scipy.sparse.coo_array((val, (row, col)), shape=(3, 3))

mdisplay(dupes)


# In[ ]: