1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393
|
"""
MCMC plotting methods.
"""
from __future__ import division, print_function
__all__ = ['plot_all', 'plot_corr', 'plot_corrmatrix',
'plot_trace', 'plot_vars', 'plot_var',
'plot_R', 'plot_logp', 'format_vars']
import math
import numpy as np
from numpy import arange, squeeze, linspace, meshgrid, vstack, inf
from scipy.stats import kde
from . import corrplot
from .formatnum import format_value
from .stats import var_stats, format_vars, save_vars
def plot_all(state, portion=1.0, figfile=None):
from pylab import figure, savefig, suptitle, rcParams
figext = '.'+rcParams.get('savefig.format', 'png')
draw = state.draw(portion=portion)
all_vstats = var_stats(draw)
figure()
plot_vars(draw, all_vstats)
if state.title:
suptitle(state.title)
print(format_vars(all_vstats))
if figfile is not None:
savefig(figfile+"-vars"+figext)
if figfile is not None:
save_vars(all_vstats, figfile+"-err.json")
figure()
plot_traces(state, portion=portion)
suptitle("Parameter history" + (" for " + state.title if state.title else ""))
if figfile is not None:
savefig(figfile+"-trace"+figext)
# Suppress R stat for now
#figure()
#plot_R(state, portion=portion)
#if state.title:
# suptitle(state.title)
#if figfile is not None:
# savefig(figfile+"-R"+format)
figure()
plot_logp(state, portion=portion)
if state.title:
suptitle(state.title)
if figfile is not None:
savefig(figfile+"-logp"+figext)
if draw.num_vars <= 25:
figure()
plot_corrmatrix(draw)
if state.title:
suptitle(state.title)
if figfile is not None:
savefig(figfile+"-corr"+figext)
def plot_vars(draw, all_vstats, **kw):
from pylab import subplot, clf
clf()
nw, nh = tile_axes(len(all_vstats))
cbar = _make_fig_colorbar(draw.logp)
for k, vstats in enumerate(all_vstats):
subplot(nw, nh, k+1)
plot_var(draw, vstats, k, cbar, **kw)
def tile_axes(n, size=None):
"""
Creates a tile for the axes which covers as much area of the graph as
possible while keeping the plot shape near the golden ratio.
"""
from pylab import gcf
if size is None:
size = gcf().get_size_inches()
figwidth, figheight = size
# Golden ratio phi is the preferred dimension
# phi = sqrt(5)/2
#
# nw, nh is the number of tiles across and down respectively
# w, h are the sizes of the tiles
#
# w,h = figwidth/nw, figheight/nh
#
# To achieve the golden ratio, set w/h to phi:
# w/h = phi => figwidth/figheight*nh/nw = phi
# => nh/nw = phi * figheight/figwidth
# Must have enough tiles:
# nh*nw > n => nw > n/nh
# => nh**2 > n * phi * figheight/figwidth
# => nh = floor(sqrt(n*phi*figheight/figwidth))
# => nw = ceil(n/nh)
phi = math.sqrt(5)/2
nh = int(math.floor(math.sqrt(n*phi*figheight/figwidth)))
if nh < 1:
nh = 1
nw = int(math.ceil(n/nh))
return nw, nh
def plot_var(draw, vstats, var, cbar, nbins=30):
values = draw.points[:, var].flatten()
_make_logp_histogram(values, draw.logp, nbins, vstats.p95_range,
draw.weights, cbar)
_decorate_histogram(vstats)
def _decorate_histogram(vstats):
import pylab
from matplotlib.transforms import blended_transform_factory as blend
l95, h95 = vstats.p95_range
l68, h68 = vstats.p68_range
# Shade things inside 1-sigma
pylab.axvspan(l68, h68, color='gold', alpha=0.5, zorder=-1)
# build transform with x=data, y=axes(0,1)
ax = pylab.gca()
transform = blend(ax.transData, ax.transAxes)
def marker(symbol, position):
if position < l95:
symbol, position, ha = '<'+symbol, l95, 'left'
elif position > h95:
symbol, position, ha = '>'+symbol, h95, 'right'
else:
symbol, position, ha = symbol, position, 'center'
pylab.text(position, 0.95, symbol, va='top', ha=ha,
transform=transform, zorder=3, color='g')
#pylab.axvline(v)
marker('|', vstats.median)
marker('E', vstats.mean)
marker('*', vstats.best)
pylab.text(0.01, 0.95, vstats.label, zorder=2,
backgroundcolor=(1, 1, 0, 0.2),
verticalalignment='top',
horizontalalignment='left',
transform=pylab.gca().transAxes)
pylab.setp([pylab.gca().get_yticklabels()], visible=False)
ticks = (l95, l68, vstats.median, h68, h95)
labels = [format_value(v, h95-l95) for v in ticks]
if len(labels[2]) > 5:
# Drop 68% values if too many digits
ticks, labels = ticks[0::2], labels[0::2]
pylab.xticks(ticks, labels)
def _make_fig_colorbar(logp):
import matplotlib as mpl
import pylab
# Option 1: min to min + 4
#vmin=-max(logp); vmax=vmin+4
# Option 1b: min to min log10(num samples)
#vmin=-max(logp); vmax=vmin+log10(len(logp))
# Option 2: full range of best 98%
snllf = pylab.sort(-logp)
vmin, vmax = snllf[0], snllf[int(0.98*(len(snllf)-1))] # robust range
# Option 3: full range
#vmin,vmax = -max(logp),-min(logp)
fig = pylab.gcf()
ax = fig.add_axes([0.60, 0.95, 0.35, 0.05])
cmap = mpl.cm.copper
# Set the colormap and norm to correspond to the data for which
# the colorbar will be used.
norm = mpl.colors.Normalize(vmin=vmin, vmax=vmax)
# ColorbarBase derives from ScalarMappable and puts a colorbar
# in a specified axes, so it has everything needed for a
# standalone colorbar. There are many more kwargs, but the
# following gives a basic continuous colorbar with ticks
# and labels.
class MinDigitsFormatter(mpl.ticker.Formatter):
def __init__(self, low, high):
self.delta = high - low
def __call__(self, x, pos=None):
return format_value(x, self.delta)
ticks = (vmin, vmax)
formatter = MinDigitsFormatter(vmin, vmax)
cb = mpl.colorbar.ColorbarBase(ax, cmap=cmap, norm=norm,
ticks=ticks, format=formatter,
orientation='horizontal')
#cb.set_ticks(ticks)
#cb.set_ticklabels(labels)
#cb.set_label('negative log likelihood')
return vmin, vmax, cmap
def _make_logp_histogram(values, logp, nbins, ci, weights, cbar):
from numpy import (ones_like, searchsorted, linspace, cumsum, diff,
argsort, array, hstack, exp)
if weights is None:
weights = ones_like(logp)
# TODO: values are being sorted to collect stats and again to plot
idx = argsort(values)
values, weights, logp = values[idx], weights[idx], logp[idx]
#open('/tmp/out','a').write("ci=%s, range=%s\n"
# % (ci,(min(values),max(values))))
edges = linspace(ci[0], ci[1], nbins+1)
idx = searchsorted(values[1:-1], edges)
weightsum = cumsum(weights)
heights = diff(weightsum[idx])/weightsum[-1] # normalized weights
import pylab
vmin, vmax, cmap = cbar
cmap_steps = linspace(vmin, vmax, cmap.N+1)
bins = [] # marginalized maximum likelihood
for h, s, e, xlo, xhi \
in zip(heights, idx[:-1], idx[1:], edges[:-1], edges[1:]):
if s == e:
continue
pv = -logp[s:e]
pidx = argsort(pv)
pw = weights[s:e][pidx]
x = array([xlo, xhi], 'd')
y = hstack((0, cumsum(pw)))
z = pv[pidx][:, None]
# centerpoint, histogram height, maximum likelihood for each bin
bins.append(((xlo+xhi)/2, y[-1], exp(vmin-z[0])))
if len(z) > cmap.N:
# downsample histogram bar according to number of colors
pidx = searchsorted(z[1:-1].flatten(), cmap_steps)
if pidx[-1] < len(z)-1:
pidx = hstack((pidx, -1))
y, z = y[pidx], z[pidx]
pylab.pcolormesh(x, y, z, vmin=vmin, vmax=vmax, cmap=cmap)
# Check for broken distribution
if not bins:
return
centers, height, maxlikelihood = array(bins).T
# Normalize maximum likelihood plot so it contains the same area as the
# histogram, unless it is really spikey, in which case make sure it has
# about the same height as the histogram.
maxlikelihood *= np.sum(height)/np.sum(maxlikelihood)
hist_peak = np.max(height)
ml_peak = np.max(maxlikelihood)
if ml_peak > hist_peak*1.3:
maxlikelihood *= hist_peak*1.3/ml_peak
pylab.plot(centers, maxlikelihood, '-g')
def _make_var_histogram(values, logp, nbins, ci, weights):
# Produce a histogram
hist, bins = np.histogram(values, bins=nbins, range=ci,
#new=True,
normed=True, weights=weights)
# Find the max likelihood for values in each bin
edges = np.searchsorted(values, bins)
histbest = [np.max(logp[edges[i]:edges[i+1]])
if edges[i] < edges[i+1] else -inf
for i in range(nbins)]
# scale to marginalized probability with peak the same height as hist
histbest = np.exp(np.asarray(histbest) - max(logp)) * np.max(hist)
import pylab
# Plot the histogram
pylab.bar(bins[:-1], hist, width=bins[1]-bins[0])
# Plot the kernel density estimate
#density = KDE1D(values)
#x = linspace(bins[0],bins[-1],100)
#pylab.plot(x, density(x), '-k')
# Plot the marginal maximum likelihood
centers = (bins[:-1]+bins[1:])/2
pylab.plot(centers, histbest, '-g')
def plot_corrmatrix(draw):
c = corrplot.Corr2d(draw.points.T, bins=50, labels=draw.labels)
c.plot()
#print "Correlation matrix\n",c.R()
class KDE1D(kde.gaussian_kde):
covariance_factor = lambda self: 2*self.silverman_factor()
class KDE2D(kde.gaussian_kde):
covariance_factor = kde.gaussian_kde.silverman_factor
def __init__(self, dataset):
kde.gaussian_kde.__init__(self, dataset.T)
def evalxy(self, x, y):
grid_x, grid_y = meshgrid(x, y)
dxy = self.evaluate(vstack([grid_x.flatten(), grid_y.flatten()]))
return dxy.reshape(grid_x.shape)
__call__ = evalxy
def plot_corr(draw, vars=(0, 1)):
from pylab import axes, setp, MaxNLocator
_, _ = vars # Make sure vars is length 2
labels = [draw.labels[v] for v in vars]
values = [draw.points[:, v] for v in vars]
# Form kernel density estimates of the parameters
xmin, xmax = min(values[0]), max(values[0])
density_x = KDE1D(values[0])
x = linspace(xmin, xmax, 100)
px = density_x(x)
density_y = KDE1D(values[1])
ymin, ymax = min(values[1]), max(values[1])
y = linspace(ymin, ymax, 100)
py = density_y(y)
nbins = 50
ax_data = axes([0.1, 0.1, 0.63, 0.63]) # x,y,w,h
#density_xy = KDE2D(values[vars])
#dxy = density_xy(x,y)*points.shape[0]
#ax_data.pcolorfast(x,y,dxy,cmap=cm.gist_earth_r) #@UndefinedVariable
ax_data.plot(values[0], values[1], 'k.', markersize=1)
ax_data.set_xlabel(labels[0])
ax_data.set_ylabel(labels[1])
ax_hist_x = axes([0.1, 0.75, 0.63, 0.2], sharex=ax_data)
ax_hist_x.hist(values[0], nbins, orientation='vertical', normed=1)
ax_hist_x.plot(x, px, 'k-')
ax_hist_x.yaxis.set_major_locator(MaxNLocator(4, prune="both"))
setp(ax_hist_x.get_xticklabels(), visible=False,)
ax_hist_y = axes([0.75, 0.1, 0.2, 0.63], sharey=ax_data)
ax_hist_y.hist(values[1], nbins, orientation='horizontal', normed=1)
ax_hist_y.plot(py, y, 'k-')
ax_hist_y.xaxis.set_major_locator(MaxNLocator(4, prune="both"))
setp(ax_hist_y.get_yticklabels(), visible=False)
def plot_traces(state, vars=None, portion=None):
from pylab import subplot, clf, subplots_adjust
if vars is None:
vars = list(range(min(state.Nvar, 6)))
clf()
nw, nh = tile_axes(len(vars))
subplots_adjust(hspace=0.0)
for k, var in enumerate(vars):
subplot(nw, nh, k+1)
plot_trace(state, var, portion)
def plot_trace(state, var=0, portion=None):
from pylab import plot, title, xlabel, ylabel
draw, points, _ = state.chains()
label = state.labels[var]
start = int((1-portion)*len(draw)) if portion else 0
plot(arange(start, len(points))*state.thinning,
squeeze(points[start:, state._good_chains, var]))
xlabel('Generation number')
ylabel(label)
def plot_R(state, portion=None):
from pylab import plot, title, legend, xlabel, ylabel
draw, R = state.R_stat()
start = int((1-portion)*len(draw)) if portion else 0
plot(arange(start, len(R)), R[start:])
title('Convergence history')
legend(['P%d' % i for i in range(1, R.shape[1]+1)])
xlabel('Generation number')
ylabel('R')
def plot_logp(state, portion=None):
from pylab import plot, title, xlabel, ylabel
draw, logp = state.logp()
start = int((1-portion)*len(draw)) if portion else 0
plot(arange(start, len(logp)), logp[start:], ',', markersize=1)
title('Log Likelihood History')
xlabel('Generation number')
ylabel('Log likelihood at x[k]')
|