"""
.. _creating_delaunay_lines:

Creating Delaunay Lines
=======================

In this example, we show how to create AequilibraE's famous Delaunay Lines, but in Python.

For more on this topic, see its `first publication <https://xl-optim.com/delaunay/>`_.

We use the Sioux Falls example once again.
"""
# %% 
# .. seealso::
#     Several functions, methods, classes and modules are used in this example:
#
#     * :func:`aequilibrae.utils.create_delaunay_network.DelaunayAnalysis`

# %%

# Imports
import pandas as pd
from uuid import uuid4
from os.path import join
import sqlite3
from tempfile import gettempdir
import matplotlib.pyplot as plt
import shapely.wkb

from aequilibrae.utils.create_example import create_example
from aequilibrae.utils.create_delaunay_network import DelaunayAnalysis

# %%

# We create an empty project on an arbitrary folder
fldr = join(gettempdir(), uuid4().hex)

project = create_example(fldr)

# %%
# Get the Delaunay Lines generation class
da = DelaunayAnalysis(project)

# Let's create the triangulation based on the zones, but we could create based on the network (centroids) too
da.create_network("zones")

# %%
# Now we get the matrix we want and create the Delaunay Lines
demand = project.matrices.get_matrix("demand_omx")
demand.computational_view(["matrix"])

# %%
# And we will call it 'delaunay_test'./ It will also be saved in the results_database.sqlite
da.assign_matrix(demand, "delaunay_test")

# %%
# we retrieve the results
conn = sqlite3.connect(join(fldr, "results_database.sqlite"))
results = pd.read_sql("Select * from delaunay_test", conn).set_index("link_id")

# %%
# Now we get the matrix we want and create the Delaunay Lines
links = pd.read_sql("Select link_id, st_asBinary(geometry) geometry from delaunay_network", project.conn)
links.geometry = links.geometry.apply(shapely.wkb.loads)
links.set_index("link_id", inplace=True)

df = links.join(results)

max_vol = df.matrix_tot.max()

for idx, lnk in df.iterrows():
    geo = lnk.geometry
    plt.plot(*geo.xy, color="blue", linewidth=4 * lnk.matrix_tot / max_vol)
plt.show()

# %%
# Close the project
project.close()