# !/usr/bin/env python
# -*- coding: utf-8 -*-
"""accuracy_plot plots the fraction of the true values that fall within a specified probability interval"""
#-----------------------------------------------------------------------------
# Boilerplate
#-----------------------------------------------------------------------------
#-----------------------------------------------------------------------------
# Imports
#-----------------------------------------------------------------------------
from . set_style import set_plot_style
[docs]
@set_plot_style
def accuracy_plot(truth=None, reals=None, mik_thresholds=None, acctype='sim', probability_increment=0.05,
figsize=None, title=None, xlabel=None, ylabel=None, stat_blk='standard',
stat_xy=(.95, .05), stat_fontsize=None, ms=5, grid=None, axis_xy=None, ax=None,
plot_style=None, custom_style=None, output_file=None, **kwargs):
"""
Accuracy plot based on probability intervals quantified using an estimation technique e.g kriging.
Currently, this plotting tool works based on using realizations to quantify the distrbution of estimation.
Two statistics block sets are available: ``'minimal'`` and the default ``'standard'``. The
statistics block can be customized to a user defined list and order. Available statistics are
as follows:
>>> ['ndat', 'nint', 'avgvar', 'mse', 'acc', 'pre', 'goo']
Please review the documentation of the :func:`gs.PlotStyle.set_style()
<pygeostat.pygeostat_parameters.PlotStyle.set_style>` and :func:`gs.export_image()
<pygeostat.plotting.export_image.export_image>` functions for details on their parameters so that
their use in this function can be understood.
Keyword Arguments:
truth: Tidy (long-form) 1D data where a single column containing the true values.
A pandas dataframe/series or numpy array can be passed
reals: Tidy (long-form) 2D data where a single column contains values from a single
realizations and each row contains the simulated values from a single truth location. “reals” has the same
number of data points as “truth”, that is , each row of “reals” corresponds to the true value in the same row of “truth”.
A pandas dataframe or numpy matrix can be passed
mik_thresholds (np.ndarray): 1D array of the z-vals ``mik_thresholds`` corresponding to the
probabilities defined in reals for each location
acctype (str): Currently ``sim`` and ``mik`` are valid. if ``mik_thresholds`` is passed the
type is assumed to be ``mik``
probability_increment (float): Probability increment used during accuracy_plot calculation
figsize (tuple): Figure size (width, height)
title (str): Title for the plot
xlabel (str): X-axis label
ylabel (str): Y-axis label
stat_blk (bool): Indicate if statistics are plotted or not
stat_xy (float tuple): X, Y coordinates of the annotated statistics in figure space. The
coordinates specify the top right corner of the text block
stat_fontsize (float): the fontsize for the statistics block. If None, based on
gs.Parameters['plotting.stat_fontsize']. If less than 1, it is the fraction of the
matplotlib.rcParams['font.size']. If greater than 1, it the absolute font size.
ms (float): Size of scatter plot markers
grid(bool): plots the major grid lines if True. Based on gs.Parameters['plotting.grid']
if None.
axis_xy (bool): converts the axis to GSLIB-style axis visibility (only left and bottom
visible) if axis_xy is True. Based on gs.Parameters['plotting.axis_xy'] if None.
ax (mpl.axis): Matplotlib axis to plot the figure
pltstyle (str): Use a predefined set of matplotlib plotting parameters as specified by
:class:`gs.GridDef <pygeostat.data.grid_definition.GridDef>`. Use ``False`` or ``None``
to turn it off
cust_style (dict): Alter some of the predefined parameters in the ``pltstyle`` selected.
output_file (str): Output figure file name and location
**kwargs: Optional permissible keyword arguments to pass to
:func:`gs.export_image() <pygeostat.plotting.export_image.export_image>`
Returns:
ax (ax): Matplotlib Axes object with the cross validation plot
Note:
The simulated values in each row of “reals” are used to calculate an empirical CDF using the standard method of
the midpoint of a histogram bin and the CDF is constructed using all points (as done in GSLIB). When "mik_thresholds" is None,
“reals” contains the simulated values from which a cdf should be computed. Otherwise the “reals” contains the distribution values,
F(mik_thresholds), for each location (each row of the data file).
Examples:
A simple call using truth and realization data: in this example, the first column of data_file is "truth"
and the rest of columns are "reals".
.. plot::
import pygeostat as gs
data_file = gs.ExampleData('accuracy_plot')
reals = data_file[list(data_file.columns[1:])].values
truth = data_file[list(data_file.columns)[0]].values
gs.accuracy_plot(truth=truth, reals=reals)
"""
import numpy as np
from ..statistics.utils import accsim, accmik
from .utils import setup_plot, get_statblk, format_plot, _set_stat_fontsize
from .export_image import export_image
import matplotlib as mpl
from .. pygeostat_parameters import Parameters
# Sanity checks based on the provided input arguments
simcheck = [(val is not None) for val in [truth, reals]]
mikcheck = [(val is not None) for val in [mik_thresholds]]
if any(simcheck) and not all(simcheck):
raise ValueError("The arguments `truth` and `reals` are required")
if any(mikcheck) and not all(simcheck):
raise ValueError("Must have `truth`, `reals` and `mik_thresholds` for MIK !")
if all(simcheck) and acctype.lower() != 'sim':
acctype = 'sim'
if all(mikcheck) and acctype == 'sim':
acctype = 'mik'
# Handle accuracy_plot-sim functionality
if acctype.lower() == 'sim':
propavg, statlist = accsim(truth, reals, probability_increment)
x = propavg['ProbInt'].values
y = propavg['FracIn'].values
elif acctype.lower() == 'mik':
propavg, statlist = accmik(truth, mik_thresholds, reals, probability_increment)
x = propavg['ProbInt'].values
y = propavg['FracIn'].values
else:
raise ValueError ('Wrong acctype was provided.')
fig, ax, cax = setup_plot(ax, cbar=False, figsize=figsize)
ax.set_aspect('equal')
# Plot the figure
for point in zip(x, y):
ax.plot((point[0], point[0]), (point[1], point[0]), color='red', zorder=0, lw=0.25)
ax.scatter(x, y, c='k', s=ms)
ax.plot((1, 0), (1, 0), zorder=0, lw=0.5, color='k')
# Configure plot
ax.axis('equal')
if xlabel is None:
xlabel = 'Probability Interval'
if ylabel is None:
ylabel = 'Fraction in Interval'
format_plot(ax, xlabel=xlabel, ylabel=ylabel, title=title,
grid=grid, xlim=(0, 1.0), ylim=(0, 1.0))
# Plot stats
if stat_blk:
statlist = {'ndat': '$n = {}$'.format(len(truth)),
'nint': '$n_p = {}$'.format(len(x)),
'avgvar': '$U = {:0.3e}$'.format(statlist['avgvar']),
'mse': '$MSE = {:0.3e}$'.format(statlist['mse']),
'acc': '$Accuracy = {:0.3e}$'.format(statlist['acc']),
'pre': '$Precision = {:0.3e}$'.format(statlist['pre']),
'goo': '$Goodness = {:0.3e}$'.format(statlist['goo'])}
statsets = {'minimal': ['nint'],
'standard': ['nint', 'avgvar', 'mse'],
"all": ['ndat', 'nint', 'avgvar', 'mse', 'acc', 'pre', 'goo']}
if len(truth) > 0:
statsets['standard'].insert(0, 'ndat')
if isinstance(stat_blk, str):
stat_blk = [stat_blk]
# Get the stat block text
txtstats, stat_xy, ha, va = get_statblk(stat_blk, statsets, statlist, stat_xy)
# Write the stat block to the plot
stat_fontsize = _set_stat_fontsize(stat_fontsize)
ax.text(stat_xy[0], stat_xy[1], txtstats, va=va, ha=ha, transform=ax.transAxes,
fontsize=mpl.rcParams['font.size'] - 0.5)
# Export figure
if output_file or ('pdfpages' in kwargs):
export_image(output_file, **kwargs)
return ax
