r"""
Effective Medium Theory Mapping
===============================

This example uses Self Consistent Effective Medium Theory to estimate the
electrical conductivity of a mixture of two phases of materials. Given
the electrical conductivity of each of the phases (:math:`\sigma_0`,
:math:`\sigma_1`), the :class:`simpeg.maps.SelfConsistentEffectiveMedium`
map takes the concentration of phase-1 (:math:`\phi_1`) and maps this to an
electrical conductivity.

This mapping is used in chapter 2 of:

Heagy, Lindsey J.(2018, in prep) *Electromagnetic methods for imaging
subsurface injections.* University of British Columbia

:author: `@lheagy <https://github.com/lheagy>`_

"""

import numpy as np
import matplotlib.pyplot as plt
from simpeg import maps
from matplotlib import rcParams

rcParams["font.size"] = 12

###############################################################################
# Conductivities
# ---------------
#
# Here we consider a mixture composed of fluid (3 S/m) and conductive
# particles which we will vary the conductivity of.
#

sigma_fluid = 3
sigma1 = np.logspace(1, 5, 5)  # look at a range of particle conductivities
phi = np.linspace(0.0, 1, 1000)  # vary the volume of particles

###############################################################################
# Construct the Mapping
# ---------------------
#
# We set the conductivity of the phase-0 material to the conductivity of the
# fluid. The mapping will then take a concentration (by volume), of phase-1
# material and compute the effective conductivity
#

scemt = maps.SelfConsistentEffectiveMedium(sigma0=sigma_fluid, sigma1=1)

###############################################################################
# Loop over a range of particle conductivities
# --------------------------------------------
#
# We loop over the values defined as `sigma1` and compute the effective
# conductivity of the mixture for each concentration in the `phi` vector
#

sige = np.zeros([phi.size, sigma1.size])

for i, s in enumerate(sigma1):
    scemt.sigma1 = s
    sige[:, i] = scemt * phi

###############################################################################
# Plot the effective conductivity
# -------------------------------
#
# The plot shows the effective conductivity of 5 difference mixtures. In all
# cases, the conductivity of the fluid, :math:`\sigma_0`, is 3 S/m. The
# conductivity of the particles is indicated in the legend
#

fig, ax = plt.subplots(1, 1, figsize=(7, 4), dpi=350)

ax.semilogy(phi, sige)
ax.grid(which="both", alpha=0.4)
ax.legend(["{:1.0e} S/m".format(s) for s in sigma1])
ax.set_xlabel(r"Volume fraction of proppant $\phi$")
ax.set_ylabel("Effective conductivity (S/m)")

plt.tight_layout()