def plot_fits(pars, samples, dolog=True, show=False, eps=None, par_labels=None):
"""
"""
import esutil as eu
import biggles
import images
biggles.configure('screen','width', 1400)
biggles.configure('screen','height', 800)
num=pars.shape[0]
ndim=pars.shape[1]
if par_labels[0]=='fracdev':
nrow=1
ncol=2
tab=biggles.Table(nrow,ncol)
plin = _plot_single(pars[:,0], samples[:,0], do_ylog=False)
plog = _plot_single(pars[:,0], samples[:,0], do_ylog=True)
plin.xlabel='fracdev'
plog.xlabel='fracdev'
tab[0,0]=plin
tab[0,1]=plog
else:
nrow,ncol = images.get_grid(ndim)
tab=biggles.Table(nrow,ncol)
for dim in xrange(ndim):
plt = _plot_single(pars[:,dim], samples[:,dim],do_ylog=True)
if par_labels is not None:
plt.xlabel=par_labels[dim]
else:
plt.xlabel=r'$P_%s$' % dim
row=(dim)/ncol
col=(dim) % ncol
tab[row,col] = plt
tab.aspect_ratio=nrow/float(ncol)
if eps:
import converter
print(eps)
d=os.path.dirname(eps)
if not os.path.exists(d):
os.makedirs(d)
tab.write_eps(eps)
converter.convert(eps, verbose=True, dpi=200)
if show:
tab.show()
def plot_fits(pars, samples, comps, dolog=True, show=False, eps=None, par_labels=None):
"""
"""
import esutil as eu
import biggles
import images
biggles.configure('screen','width', 1400)
biggles.configure('screen','height', 800)
num=pars.shape[0]
ndim=pars.shape[1]
nrow,ncol = images.get_grid(ndim)
tab=biggles.Table(nrow,ncol)
for dim in xrange(ndim):
plt = _plot_single(pars[:,dim], samples[:,dim], comps, do_ylog=True)
if par_labels is not None:
plt.xlabel=par_labels[dim]
else:
plt.xlabel=r'$P_%s$' % dim
row=(dim)/ncol
col=(dim) % ncol
tab[row,col] = plt
tab.aspect_ratio=nrow/float(ncol)
if eps:
import converter
print(eps)
d=os.path.dirname(eps)
if not os.path.exists(d):
os.makedirs(d)
tab.write_eps(eps)
converter.convert(eps, verbose=True, dpi=200)
if show:
tab.show()
def plot_results_separate(trials, **keys):
"""
Plot the points and histograms of trials from an MCMC chain.
"""
import biggles
import esutil
import images
npars = trials.shape[1]
fontsize_min = keys.get("fontsize_min", 1)
biggles.configure("default", "fontsize_min", fontsize_min)
weights = keys.get("weights", None)
binfac = keys.get("binfac", 0.2)
names = keys.get("names", None)
show = keys.get("show", True)
ptypes = keys.get("ptypes", ["linear"] * npars)
means, cov = extract_stats(trials, weights=weights)
errs = sqrt(diag(cov))
nrows, ncols = images.get_grid(npars)
burn_plt = biggles.Table(nrows, ncols)
hist_plt = biggles.Table(nrows, ncols)
ind = numpy.arange(trials.shape[0])
prow = 0
pcol = 0
for i in xrange(npars):
prow = i / ncols
pcol = i % ncols
if names is not None:
name = names[i]
else:
name = r"$p_{%d}$" % i
lab = r"$<%s> = %0.4g \pm %0.4g$" % (name, means[i], errs[i])
plab = biggles.PlotLabel(0.1, 0.8, lab, halign="left", color="blue")
# steps
burn_plot_i = esutil.plotting.bscatter(ind, trials[:, i], type="solid", xlabel="step", ylabel=name, show=False)
burn_plot_i.add(plab)
burn_plt[prow, pcol] = burn_plot_i
# hist
vals = trials[:, i]
bsize = binfac * errs[i]
hdict = esutil.stat.histogram(vals, binsize=bsize, weights=weights, more=True)
if weights is not None:
hist = hdict["whist"]
hplot = biggles.Curve(hdict["center"], hdict["whist"])
else:
hist = hdict["hist"]
hplot = biggles.Histogram(hdict["hist"], x0=hdict["low"][0], binsize=bsize)
plti = biggles.FramedPlot()
plti.xlabel = name
hmax = hist.max()
plti.yrange = [-0.05 * hmax, 1.2 * hmax]
plti.add(hplot)
plti.add(plab)
hist_plt[prow, pcol] = plti
burn_plt.title = keys.get("title", None)
hist_plt.title = keys.get("title", None)
if show:
burn_plt.show()
hist_plt.show()
return burn_plt, hist_plt
def plot_results_separate(trials, **keys):
"""
Plot the points and histograms of trials from an MCMC chain.
"""
import biggles
import esutil
import images
npars=trials.shape[1]
fontsize_min=keys.get('fontsize_min',1)
biggles.configure( 'default', 'fontsize_min', fontsize_min)
weights=keys.get('weights',None)
binfac=keys.get('binfac',0.2)
names=keys.get('names',None)
show=keys.get('show',True)
ptypes=keys.get('ptypes',['linear']*npars)
means,cov = extract_stats(trials,weights=weights)
errs=sqrt(diag(cov))
nrows, ncols =images.get_grid(npars)
burn_plt=biggles.Table(nrows, ncols)
hist_plt=biggles.Table(nrows, ncols)
ind = numpy.arange(trials.shape[0])
prow=0
pcol=0
for i in xrange(npars):
prow=i/ncols
pcol=i % ncols
if names is not None:
name=names[i]
else:
name=r'$p_{%d}$' % i
lab = r'$<%s> = %0.4g \pm %0.4g$' % (name,means[i],errs[i])
plab = biggles.PlotLabel(0.1,0.8,lab,
halign='left',
color='blue')
# steps
burn_plot_i = esutil.plotting.bscatter(ind,trials[:,i],
type='solid',
xlabel='step',
ylabel=name,
show=False)
burn_plot_i.add(plab)
burn_plt[prow,pcol] = burn_plot_i
# hist
vals=trials[:,i]
bsize = binfac*errs[i]
hdict = esutil.stat.histogram(vals,
binsize=bsize,
weights=weights,
more=True)
if weights is not None:
hist=hdict['whist']
hplot = biggles.Curve(hdict['center'],
hdict['whist'])
else:
hist=hdict['hist']
hplot = biggles.Histogram(hdict['hist'],
x0=hdict['low'][0],
binsize=bsize)
plti=biggles.FramedPlot()
plti.xlabel=name
hmax=hist.max()
plti.yrange=[-0.05*hmax, 1.2*hmax]
plti.add(hplot)
plti.add(plab)
hist_plt[prow,pcol]=plti
burn_plt.title=keys.get('title',None)
hist_plt.title=keys.get('title',None)
if show:
burn_plt.show()
hist_plt.show()
return burn_plt, hist_plt
