def errorpage(errval, emailaddress, outdir):
# the string containing the webpage
htmlpage = """
<!DOCTYPE HTML>
<html>
<head>
<title>The Online MCMC: Error page</title>
<link rel="stylesheet" type="text/css" href="../../simple.css"/>
</head>
<body>
<div id="page-wrap">
<!-- include header file -->
<?php include('../../header.inc'); ?>
<p>{errormessage}</p>
<!-- include footer file -->
<?php include('../../footer.inc'); ?>
</div>
</body>
"""
# the output php file
errfile = 'index.php'
fp = open(errfile, 'w')
fp.write(htmlpage.format(errormessage=errormessages[errval]))
fp.close()
# email the page
emailresponse(emailaddress, outdir, runerror=True)
def postprocessing(corrcoef, postsamples, variables, abscissa, abscissaname,
data, email, outdir, evidence):
# import the corner plot code
import corner
# import matplotlib
from matplotlib import pyplot as pl
# the format text to include in the page
fm = {}
# the string containing the webpage
htmlpage = """
<!DOCTYPE HTML>
<html>
<head>
<!-- Theme Made By www.w3schools.com - No Copyright -->
<meta name="author" content="Matthew Pitkin">
<meta name="description" content="The Online MCMC">
<meta name="keywords" content="MCMC, Markov chain Monte Carlo, Bayesian, emcee, python, data analysis, probability">
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/css/bootstrap.css">
<script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js"></script>
<!-- Include theme font -->
<link href="https://fonts.googleapis.com/css?family=Montserrat" rel="stylesheet">
<!-- Include jQuery -->
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.2.1/jquery.min.js"></script>
<!-- custom CSS file -->
<link rel="stylesheet" type="text/css" href="../../simple.css"/>
<!-- include MathJax -->
<script type="text/javascript" src="https://cdn.mathjax.org/mathjax/latest/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
<!-- Formatting for text on website -->
<style>body{{color:dimgrey}}</style>
<!-- Include script to create tabs - https://www.w3schools.com/bootstrap/bootstrap_tabs_pills.asp -->
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.3.1/jquery.min.js"></script>
<script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js"></script>
<!-- Make sure text is readable with previous css formatting -->
<style>body{{color:dimgrey}}</style>
<title>The Online MCMC: Results page</title>
</head>
<body>
<!-- Navbar -->
<nav class="navbar navbar-default">
<div class="container">
<div class="navbar-header">
<button type="button" class="navbar-toggle" data-toggle="collapse" data-target="#myNavbar">
<span class="icon-bar"></span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
</button>
<a class="navbar-brand" href="#">THE ONLINE MCMC</a>
</div>
<div class="collapse navbar-collapse" id="myNavbar">
<ul class="nav navbar-nav navbar-right">
<li><a href="#margpos">MARGINALISED POSTERIORS</a></li>
<li><a href="#bestfit">BEST FIT VALUES & DISTRIBUTION</a></li>
<li><a href="#links">LINKS</a></li>
</ul>
</div>
</div>
</nav>
<div class="container-top bg-1 text-center">
<h2 class="title">RESULTS</h2>
<h3>Your results have been generated and are displayed below.</h3>
</div>
<div class="container">
<ul class="nav nav-pills">
<li class="active"><a data-toggle="pill" href="#margpos">MARGINALISED POSTERIORS</a></li>
<li><a data-toggle="pill" href="#bestfit">BEST FIT VALUES & DISTRIBUTION</a></li>
<li><a data-toggle="pill" href="#links">LINKS</a></li>
</ul>
<div class="tab-content">
<div id="margpos" class="tab-pane fade in active">
<h3 class="text-center" id="instructions">MARGINALISED POSTERIORS</h3>
<p>
The diagonal plots show the <a style="color: #BD5D38" href="https://en.wikipedia.org/wiki/Marginal_distribution">marginal</a> <a style="color: #BD5D38" href="https://en.wikipedia.org/wiki/Posterior_probability">posterior probability</a> distributions for each of your fitted parameters. The off-diagonal plots show 1 and 2σ probability contours for the <a style="color: #BD5D38" href="https://en.wikipedia.org/wiki/Joint_probability_distribution">joint</a> marginal posterior probability distributions of pairs of parameters. This has been produced with <a style="color: #BD5D38" href="https://github.com/dfm/corner.py">corner.py</a>.
</p>
{posteriorfig}
</div>
<div id="bestfit" class="tab-pane fade">
<h3 class="text-center" id="functions">BEST FIT VALUES</h3>
The <a href="https://en.wikipedia.org/wiki/Mean">mean</a> and <a href="https://en.wikipedia.org/wiki/Median">median</a> of the probability distributions for each parameter are displayed in the table below. Also given are the <a href="https://en.wikipedia.org/wiki/Standard_deviation">standard deviation</a> of the distributions and minimal 68% and 95% <a href="https://en.wikipedia.org/wiki/Credible_interval">credible intervals</a> (CI).
<div>
{resultstable}
</div>
<h3>Correlation coefficient matrix</h3>
The <a href="https://en.wikipedia.org/wiki/Covariance_matrix#Correlation_matrix">correlation coefficients</a> between each of the fitted parameters.
<div>
{corrcoeftable}
</div>
{evidencevalue}
<div class="container-fluid bg-1 text-left">
<h3 class="text-center" id="instructions">BEST FIT MODEL DISTRIBUTION</h3>
<p>
This plot shows the distribution of 100 models drawn randomly from the posterior distribution. The best fitting models will be clustered over each other.
</p>
{bestfitfig}
</div>
</div>
<div id="links" class="tab-pane fade">
<h2 class="text-center">LINKS</h2>
<h3>Code links</h3>
The <a href="https://www.python.org/">python</a> files for running the MCMC are provided below. These use the <a href="https://pypi.org/project/bilby/">bilby</a> python module.
<ul>
<li><a href="pyfile.py"><code>pyfile.py</code></a> - the python file used to run the MCMC</li>
<li><a href="mymodel.py"><code>mymodel.py</code></a> - the python model function</li>
</ul>
<h3>Data links</h3>
<ul>
<li><a href="posterior_samples.txt.gz"><code>posterior_samples.txt.gz</code></a> - a gzipped tarball containing the posterior samples</li>
<li><a href="variables.txt"><code>variables.txt</code></a> - the variables in the order of the posterior file</li>
</ul>
</div>
</div>
</div>
<footer class="container-fluid bg-2 text-center">
<p class="footer"> © Matthew Pitkin (2015), Catriona Marr (2018), Francis Webb (2019). The code for this site is licensed under the <a style="color: #BD5D38" href="http://opensource.org/licenses/MIT">MIT license</a>. It is available on <a style="color: #BD5D38" href="https://github.com/mattpitkin/theonlinemcmc">github</a> and <a style="color: #BD5D38" href="https://bitbucket.org/mattpitkin/theonlinemcmc">bitbucket</a>.<br>This site is kindly hosted by the <a style="color: #BD5D38" href="http://www.gla.ac.uk/schools/physics/">School of Physics & Astronomy</a> at the <a style="color: #BD5D38" href="http://www.gla.ac.uk/">University of Glasgow</a>. They bear no reponsibility for the content of this site or the results that are produced.
</p>
<!-- include social media sharing -->
<?php
$shareurl = "{outdir}";
include('../../social.inc');
?>
</div>
</body>
"""
fm['outdir'] = outdir
varnames = variables.split(',')
nvars = len(varnames)
# convert any Greek alphabet variable names into LaTeX tags (prefix with \)
for i, var in enumerate(varnames):
if var in greekletters:
varnames[i] = '\\' + varnames[i]
# if sigma has been fit for the Gaussian likelihood change the variable name
if "sigma_gauss" in var:
varnames[i] = '\\sigma_{\mathrm{gauss}}' # convert to LaTeX
# create triangle plot
fm['posteriorfig'] = '<img class="center-block bg-3" src="outdir/label_corner.png" width="60%">'
# get the 68% and 95% credible intervals
inter68 = []
inter95 = []
for i in range(nvars):
inter68.append(credible_interval(postsamples[:, i], 0.68))
inter95.append(credible_interval(postsamples[:, i], 0.95))
# output the results table
resultstable = "<table class=\"table table-striped table-hover\">\n<tr><th>Variable</th><th>Mean</th><th>Median</th><th>σ</th><th>68% CI</th><th>95% CI</th></tr>\n"
for i in range(nvars):
resstrs = ['\(' + varnames[i] + '\)'
] # the \( \) are the MathJax equation delimiters
meanv = np.mean(postsamples[:, i])
if np.fabs(meanv) > 1e3 or np.fabs(meanv) < 1e-2:
meanstr = exp_str(meanv)
else:
meanstr = '%.1f' % meanv
resstrs.append(meanstr)
medianv = np.median(postsamples[:, i])
if np.fabs(medianv) > 1e3 or np.fabs(medianv) < 1e-2:
medianstr = exp_str(medianv)
else:
medianstr = '%.1f' % medianv
resstrs.append(medianstr)
#modev = postsamples[np.argmax(postsamples[:,-1]),i]
#if np.fabs(modev) > 1e3 or np.fabs(modev) < 1e-2:
#modestr = exp_str(modev)
#else:
#modestr = '%.1f' % modev
#resstrs.append(modestr)
sigmav = np.std(postsamples[:, i])
if np.fabs(sigmav) > 1e3 or np.fabs(sigmav) < 1e-2:
sigmastr = exp_str(sigmav)
else:
sigmastr = '%.1f' % sigmav
resstrs.append(sigmastr)
ci68str = "[{0}, {1}]"
if np.fabs(inter68[i][0]) > 1e3 or np.fabs(inter68[i][0]) < 1e-2:
cis1 = exp_str(inter68[i][0])
else:
cis1 = '%.1f' % inter68[i][0]
if np.fabs(inter68[i][1]) > 1e3 or np.fabs(inter68[i][1]) < 1e-2:
cis2 = exp_str(inter68[i][1])
else:
cis2 = '%.1f' % inter68[i][1]
resstrs.append(ci68str.format(cis1, cis2))
ci95str = "[{0}, {1}]"
if np.fabs(inter95[i][0]) > 1e3 or np.fabs(inter95[i][0]) < 1e-2:
cis1 = exp_str(inter95[i][0])
else:
cis1 = '%.1f' % inter95[i][0]
if np.fabs(inter95[i][1]) > 1e3 or np.fabs(inter95[i][1]) < 1e-2:
cis2 = exp_str(inter95[i][1])
else:
cis2 = '%.1f' % inter95[i][1]
resstrs.append(ci95str.format(cis1, cis2))
resultstable += "<tr><td>{0}</td><td>{1}</td><td>{2}</td><td>{3}</td><td>{4}</td><td>{5}</td></tr>".format(
*resstrs)
resultstable += "</table>\n"
fm['resultstable'] = resultstable
# get the correlation coefficient matrix
# corrcoef = np.corrcoef(postsamples[:,:nvars].T)
corrcoeftable = "<table class=\"table table-striped table-hover\"><th></th>"
for i in range(nvars):
corrcoeftable += "<td>%s</td>" % ('\(' + varnames[i] + '\)')
corrcoeftable += "</tr>\n"
for i in range(nvars):
corrcoeftable += "<tr><td>%s</td>" % ('\(' + varnames[i] + '\)')
for j in range(nvars):
# check if only 1d corrcoef array
if nvars == 1:
corrcoeftable += "<td>%.2f</td>" % corrcoef
else:
corrcoeftable += "<td>%.2f</td>" % corrcoef[i, j]
corrcoeftable += "</tr>\n"
corrcoeftable += "</table>\n"
fm['corrcoeftable'] = corrcoeftable
if np.isnan(evidence) == True:
fm['evidencevalue'] = "" # Not using nestling sampler - no evidence value to display
else:
evidencevalue = "<div id=\"lnevd\" class=\"container-fluid bg-2 text-left\">"
evidencevalue += "<h3 class=\"text-center\" id=\"functions\">LOG EVIDENCE VALUE</h3>"
evidencevalue += "The nested sampler used gave a log evidence value of: <b>%.4f </b></div>" % evidence
fm['evidencevalue'] = evidencevalue
# create plot of data along with distribution of best fit models
fm['bestfitfig'] = '<img class="center-block bg-3" src="outdir/label_plot_with_data.png" width="60%">'
# output page
ppfile = 'index.php'
fp = open(ppfile, 'w')
fp.write(htmlpage.format(**fm))
fp.close()
# email the page
emailresponse(email, outdir)
def errorpage(erroroutput, errval, emailaddress, outdir):
# the string containing the webpage
htmlpage = """
<!DOCTYPE HTML>
<html>
<head>
<!-- Theme Made By www.w3schools.com - No Copyright -->
<meta name="author" content="Matthew Pitkin">
<meta name="description" content="The Online MCMC">
<meta name="keywords" content="MCMC, Markov chain Monte Carlo, Bayesian, emcee, python, data analysis, probability">
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/css/bootstrap.css">
<script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js"></script>
<!-- Include theme font -->
<link href="https://fonts.googleapis.com/css?family=Montserrat" rel="stylesheet">
<!-- Include jQuery -->
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.2.1/jquery.min.js"></script>
<!-- custom CSS file -->
<link rel="stylesheet" type="text/css" href="simple.css"/><title>The Online MCMC: Error page</title>
<!-- Include jQuery -->
<script src="http://ajax.googleapis.com/ajax/libs/jquery/1.11.3/jquery.min.js"></script>
<!-- custom CSS file -->
<link rel="stylesheet" type="text/css" href="../../simple.css"/>
<title>The Online MCMC: Error page</title>
<body>
<!-- Navbar -->
<nav class="navbar navbar-default">
<div class="container">
<div class="navbar-header">
<button type="button" class="navbar-toggle" data-toggle="collapse" data-target="#myNavbar">
<span class="icon-bar"></span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
</button>
<a class="navbar-brand" href="/">THE ONLINE MCMC</a>
</div>
</div>
</nav>
<div class="container-top bg-1 text-center">
<h3 class="title">ERROR</h3>
<p>
<br><br>{errormessage}<br>
<br><code>{erroroutput}</code><br><br>
</p>
<br>
<br>
<li><a href="pyfile.py"><code>pyfile.py</code></a> - the python file used to run the MCMC</li>
<li><a href="mymodel.py"><code>mymodel.py</code></a> - the python model function</li>
</div>
<!-- include footer -->
<?php
$shareurl = "http://www.theonlinemcmc.com";
include('../../footer.inc');
?>
<!-- include Social Media sharing file -->
<?php
$shareurl = "http://www.theonlinemcmc.com";
include('../../social.inc');
?>
</div>
</body>
"""
# the output php file
errfile = 'index.php'
fp = open(errfile, 'w')
fp.write(
htmlpage.format(errormessage=errormessages[errval],
erroroutput=erroroutput))
fp.close()
# email the page
emailresponse(emailaddress, outdir, runerror=True)
def errorpage(erroroutput, errval, emailaddress, outdir):
# the string containing the webpage
htmlpage = """
<!DOCTYPE HTML>
<html>
<head>
<!-- Theme Made By www.w3schools.com - No Copyright -->
<meta name="author" content="Matthew Pitkin">
<meta name="description" content="The Online MCMC">
<meta name="keywords" content="MCMC, Markov chain Monte Carlo, Bayesian, emcee, python, data analysis, probability">
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/css/bootstrap.css">
<script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js"></script>
<!-- Include theme font -->
<link href="https://fonts.googleapis.com/css?family=Montserrat" rel="stylesheet">
<!-- Include jQuery -->
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.2.1/jquery.min.js"></script>
<!-- custom CSS file -->
<link rel="stylesheet" type="text/css" href="simple.css"/><title>The Online MCMC: Error page</title>
<!-- Include jQuery -->
<script src="http://ajax.googleapis.com/ajax/libs/jquery/1.11.3/jquery.min.js"></script>
<!-- custom CSS file -->
<link rel="stylesheet" type="text/css" href="../../simple.css"/>
<title>The Online MCMC: Error page</title>
<body>
<!-- Navbar -->
<nav class="navbar navbar-default">
<div class="container">
<div class="navbar-header">
<button type="button" class="navbar-toggle" data-toggle="collapse" data-target="#myNavbar">
<span class="icon-bar"></span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
</button>
<a class="navbar-brand" href="/">THE ONLINE MCMC</a>
</div>
</div>
</nav>
<div class="container-top bg-1 text-center">
<h3 class="title">ERROR</h3>
<p>
<br><br>{errormessage}<br>
<br><code>{erroroutput}</code><br><br>
</p>
<br>
<br>
<li><a href="pyfile.py"><code>pyfile.py</code></a> - the python file used to run the MCMC</li>
<li><a href="mymodel.py"><code>mymodel.py</code></a> - the python model function</li>
</div>
<footer class="container-fluid bg-2 text-center">
<p class="footer"> <strong> © Matthew Pitkin (2015). </strong>The code for this site is licensed under the <a style="color: #BD5D38" href="http://opensource.org/licenses/MIT">MIT license</a>. It is available on <a style="color: #BD5D38" href="https://github.com/mattpitkin/theonlinemcmc">github</a> and <a style="color: #BD5D38" href="https://bitbucket.org/mattpitkin/theonlinemcmc">bitbucket</a>.<br>This site is kindly hosted by the <a style="color: #BD5D38" href="http://www.gla.ac.uk/schools/physics/">School of Physics & Astronomy</a> at the <a style="color: #BD5D38" href="http://www.gla.ac.uk/">University of Glasgow</a>. They bear no reponsibility for the content of this site or the results that are produced.
</p>
<!-- include Social Media sharing file -->
<?php
$shareurl = "http://www.theonlinemcmc.com";
include('../../social.inc');
?>
</div>
</body>
"""
# the output php file
errfile = 'index.php'
fp = open(errfile, 'w')
fp.write(htmlpage.format(errormessage=errormessages[errval],erroroutput=erroroutput))
fp.close()
# email the page
emailresponse(emailaddress, outdir, runerror=True)
def postprocessing(postsamples, variables, abscissa, abscissaname, data, email, outdir):
# import the corner plot code
import corner
# import matplotlib
from matplotlib import pyplot as pl
# the format text to include in the page
fm = {}
# the string containing the webpage
htmlpage = """
<!DOCTYPE HTML>
<html>
<head>
<title>The Online MCMC: Results page</title>
<link rel="stylesheet" type="text/css" href="../../simple.css"/>
<script type="text/javascript"
src="https://cdn.mathjax.org/mathjax/latest/MathJax.js?config=TeX-AMS-MML_HTMLorMML">
</script>
</head>
<body>
<div id="page-wrap">
<!-- include header file -->
<?php include('../../header.inc'); ?>
<!-- include social media sharing -->
<?php
$shareurl = "{outdir}";
include('../../social.inc');
?>
<h2>Marginalised posteriors</h2>
The diagonal plots show the <a href="https://en.wikipedia.org/wiki/Marginal_distribution">marginal</a>
<a href="https://en.wikipedia.org/wiki/Posterior_probability">posterior probability</a> distributions for each of your
fitted parameters. The off-diagonal plots show 1 and 2σ probability contours for the
<a href="https://en.wikipedia.org/wiki/Joint_probability_distribution">joint</a> marginal posterior probability distributions of pairs of parameters. This has been produced with <a href="https://github.com/dfm/corner.py">corner.py</a>.
{posteriorfig}
<h2>Best fit values</h2>
The <a href="https://en.wikipedia.org/wiki/Mean">mean</a>,
<a href="https://en.wikipedia.org/wiki/Median">median</a> and
<a href="https://en.wikipedia.org/wiki/Mode_(statistics)">mode</a>
of the probability distributions for each parameter. Also give are the
<a href="https://en.wikipedia.org/wiki/Standard_deviation">standard deviation</a>
of the distributions and minimal 68% and 95%
<a href="https://en.wikipedia.org/wiki/Credible_interval">credible intervals</a> (CI).
<div>
{resultstable}
</div>
<h3>Correlation coefficient matrix</h3>
The <a href="https://en.wikipedia.org/wiki/Covariance_matrix#Correlation_matrix">correlation coefficients</a>
between each of the fitted parameters.
<div>
{corrcoeftable}
</div>
<h2>Best fit model distribution</h2>
This plot shows the distribution of 100 models drawn randomly from the posterior distribution.
The best fitting models will be clustered over each other.
{bestfitfig}
<h2>Code links</h2>
The <a href="https://www.python.org/">python</a> files for running the MCMC are provided below.
These use the <a href="http://dan.iel.fm/emcee/current/">emcee</a> python module.
<ul>
<li><a href="pyfile.py"><code>pyfile.py</code></a> - the python file used to run the MCMC</li>
<li><a href="mymodel.py"><code>mymodel.py</code></a> - the python model function</li>
</ul>
<h2>Data links</h2>
<ul>
<li><a href="posterior_samples.txt.gz"><code>posterior_samples.txt.gz</code></a> - a gzipped tarball
containing the posterior samples</li>
<li><a href="variables.txt"><code>variables.txt</code></a> - the variables in the order of the posterior file</li>
</ul>
<!-- include footer file -->
<?php include('../../footer.inc'); ?>
</div>
</body>
"""
fm['outdir'] = outdir
varnames = variables.split(',')
# convert any Greek alphabet variable names into LaTeX tags (prefix with \)
for i, var in enumerate(varnames):
if var in greekletters:
varnames[i] = '\\'+varnames[i]
# if sigma has been fit for the Gaussian likelihood change the variable name
if "sigma_gauss" in var:
varnames[i] = '\\sigma_{\mathrm{gauss}}' # convert to LaTeX
# create triangle plot
labels = ['$%s$' % var for var in varnames] # LaTeX labels
nvars = len(varnames)
levels = 1.-np.exp(-0.5*np.array([1., 2.])**2) # plot 1 and 2 sigma contours
fig = corner.corner(postsamples[:,:nvars], labels=labels, levels=levels, quantiles=[0.16, 0.5, 0.84], bins=30)
# output the figure
postfigfile = 'posterior_plots.png'
fig.savefig(postfigfile, transparent=True) # transparent background
fm['posteriorfig'] = '<img src="' + postfigfile + '" >'
# get the 68% and 95% credible intervals
inter68 = []
inter95 = []
for i in range(nvars):
inter68.append(credible_interval(postsamples[:,i], 0.68))
inter95.append(credible_interval(postsamples[:,i], 0.95))
# output the results table
resultstable = "<table class=\"table table-striped table-hover\">\n<tr><th>Variable</th><th>Mean</th><th>Median</th><th>Mode</th><th>σ</th><th>68% CI</th><th>95% CI</th></tr>\n"
for i in range(nvars):
resstrs = ['\('+varnames[i]+'\)'] # the \( \) are the MathJax equation delimiters
meanv = np.mean(postsamples[:,i])
if np.fabs(meanv) > 1e3 or np.fabs(meanv) < 1e-2:
meanstr = exp_str(meanv)
else:
meanstr = '%.1f' % meanv
resstrs.append(meanstr)
medianv = np.median(postsamples[:,i])
if np.fabs(medianv) > 1e3 or np.fabs(medianv) < 1e-2:
medianstr = exp_str(medianv)
else:
medianstr = '%.1f' % medianv
resstrs.append(medianstr)
modev = postsamples[np.argmax(postsamples[:,-1]),i]
if np.fabs(modev) > 1e3 or np.fabs(modev) < 1e-2:
modestr = exp_str(modev)
else:
modestr = '%.1f' % modev
resstrs.append(modestr)
sigmav = np.std(postsamples[:,i])
if np.fabs(sigmav) > 1e3 or np.fabs(sigmav) < 1e-2:
sigmastr = exp_str(sigmav)
else:
sigmastr = '%.1f' % sigmav
resstrs.append(sigmastr)
ci68str = "[{0}, {1}]"
if np.fabs(inter68[i][0]) > 1e3 or np.fabs(inter68[i][0]) < 1e-2:
cis1 = exp_str(inter68[i][0])
else:
cis1 = '%.1f' % inter68[i][0]
if np.fabs(inter68[i][1]) > 1e3 or np.fabs(inter68[i][1]) < 1e-2:
cis2 = exp_str(inter68[i][1])
else:
cis2 = '%.1f' % inter68[i][1]
resstrs.append(ci68str.format(cis1, cis2))
ci95str = "[{0}, {1}]"
if np.fabs(inter95[i][0]) > 1e3 or np.fabs(inter95[i][0]) < 1e-2:
cis1 = exp_str(inter95[i][0])
else:
cis1 = '%.1f' % inter95[i][0]
if np.fabs(inter95[i][1]) > 1e3 or np.fabs(inter95[i][1]) < 1e-2:
cis2 = exp_str(inter95[i][1])
else:
cis2 = '%.1f' % inter95[i][1]
resstrs.append(ci95str.format(cis1, cis2))
resultstable += "<tr><td>{0}</td><td>{1}</td><td>{2}</td><td>{3}</td><td>{4}</td><td>{5}</td><td>{6}</td></tr>".format(*resstrs)
resultstable += "</table>\n"
fm['resultstable'] = resultstable
# get the correlation coefficient matrix
corrcoef = np.corrcoef(postsamples[:,:nvars].T)
corrcoeftable = "<table class=\"table table-striped table-hover\"><th></th>"
for i in range(nvars):
corrcoeftable += "<td>%s</td>" % ('\('+varnames[i]+'\)')
corrcoeftable += "</tr>\n"
for i in range(nvars):
corrcoeftable += "<tr><td>%s</td>" % ('\('+varnames[i]+'\)')
for j in range(nvars):
# check if only 1d corrcoef array
if nvars == 1:
corrcoeftable += "<td>%.2f</td>" % corrcoef
else:
corrcoeftable += "<td>%.2f</td>" % corrcoef[i,j]
corrcoeftable += "</tr>\n"
corrcoeftable += "</table>\n"
fm['corrcoeftable'] = corrcoeftable
# create plot of data along with distribution of best fit models
from mymodel import mymodel
# set some plot defaults
pl.rc('text', usetex=True)
pl.rc('font', family='serif')
pl.rc('font', size=14)
fig2 = pl.figure(figsize=(7,5), dpi=200)
pl.plot(abscissa, data, 'ko', ms=6, label='Data')
varidxs = range(nvars)
if 'sigma_gauss' in variables:
sigmaidx = (variables.split(',')).index('sigma_gauss')
varidxs.remove(sigmaidx)
varidxs = np.array(varidxs)
randidxs = np.random.permutation(postsamples.shape[0])[0:100]
# overplot models for 100 random draws from the posterior
for i in range(100):
thesevars = postsamples[randidxs[i], varidxs].tolist()
thesevars.append(abscissa)
thismodel = mymodel(*thesevars) # unpack list as arguments of model function
pl.plot(abscissa, thismodel, '-', color='mediumblue', lw=4, alpha=0.05)
pl.legend(loc='best', numpoints=1)
if abscissaname in greekletters:
abscissaname = '\\'+abscissaname
pl.xlabel("$"+abscissaname+"$")
modelplot = 'model_plot.png'
# later try converting to d3 figure using http://mpld3.github.io/ (e.g. import mpld3; mpld3.save_html(fig2))
fig2.savefig(modelplot, transparent=True)
fm['bestfitfig'] = '<img src="' + modelplot + '" width="100%">'
# output page
ppfile = 'index.php'
fp = open(ppfile, 'w')
fp.write(htmlpage.format(**fm))
fp.close()
# email the page
emailresponse(email, outdir)
