In [1]:
%matplotlib inline
import numpy as np
from scotchcorner import scotchcorner as sc
In [2]:
# a 2d joint distribution plot
# number of parameters
ndims = 2
# ratio of joint plots to histogram plots
ratio = 3
datatitle = '$x$'
z = np.random.randn(2000,ndims)
z[:,0] = np.abs(z[:,0])*1e-6
z[:,1] = z[:,1]
limits = [(0., None), ()]
#limits = None
p = sc(z, hist_kwargs={'histtype': 'stepfilled', 'color': 'blue', 'linewidth': 1}, showlims='joint',
showpoints=True, showcontours=True, show_level_labels=False,
contour_kwargs={'colors': 'blue'}, contour_limits=limits, limits=limits, datatitle=datatitle,
thinpoints=2.0)
In [4]:
# number of parameters
ndims = 4
# ratio of joint plots to histogram plots
ratio = 3
x = np.zeros((2000,ndims))
x[:,0:2] = np.random.randn(2000,2)
x[0:500,2] = -1. + 0.25*np.random.randn(500)
x[500:2000,2] = 3. + 1.25*np.random.randn(1500)
x[0:1000, 3] = 1.*np.random.randn(1000)
x[1000:2000, 3] = 3.+0.75*np.random.randn(1000)
x[:,0] = x[:,0]*1e-6
x[:,1] = x[:,1]*1e-8
x[:,3] = x[:,3]*1e-10
showlims = 'both'
labels = ['$\\textrm{b}$', '$x$', '$\phi$', '$y$ [$\\textrm{s}$]']
histops = {'histtype': 'stepfilled', 'color': 'darkslategrey', 'edgecolor': 'black', 'linewidth': 1.5}
limits = [(-5., 5.), (None, None), (None, None), (None, None)]
# create plot
p2 = sc(x, labels=labels, truths=[0., 1.52e-8, 0., 0.], showlims=showlims,
hist_kwargs=histops, datatitle='Data 1', figsize=(10,10), limits=limits)
z = np.random.randn(2000,ndims)
z[:,0] = z[:,0]*1e-6
z[:,1] = z[:,1]*1e-8
z[:,3] = z[:,3]*1e-10
p2.add_data(z, hist_kwargs={'histtype': 'step', 'color': 'blue', 'linewidth': 1}, datatitle='Data 2',
showcontours=True, contour_kwargs={'colors': 'blue'})