SF.net SVN: matplotlib: [4029] trunk/py4science/examples

[<jd...@us...>]

Revision: 4029
          http://matplotlib.svn.sourceforge.net/matplotlib/?rev=4029&view=rev
Author:   jdh2358
Date:     2007-10-26 13:15:51 -0700 (Fri, 26 Oct 2007)

Log Message:
-----------
added stats distributions examples

Modified Paths:
--------------
    trunk/py4science/examples/stats_descriptives.py
    trunk/py4science/examples/stats_distributions.py

Modified: trunk/py4science/examples/stats_descriptives.py
===================================================================
--- trunk/py4science/examples/stats_descriptives.py	2007-10-26 20:15:42 UTC (rev 4028)
+++ trunk/py4science/examples/stats_descriptives.py	2007-10-26 20:15:51 UTC (rev 4029)
@@ -76,21 +76,27 @@
         c = C()
         N = 5
         fig = c.fig = figfunc()
+	fig.subplots_adjust(hspace=0.3)
         ax = c.ax1 = fig.add_subplot(N,1,1)
         c.plot = ax.plot(data, fmt)
+	ax.set_ylabel('data')
 
         ax = c.ax2 = fig.add_subplot(N,1,2)
         c.hist = ax.hist(data, bins)
+	ax.set_ylabel('hist')
 
-
         ax = c.ax3 = fig.add_subplot(N,1,3)
-        c.acorr = ax.acorr(data, detrend=detrend, usevlines=True, maxlags=maxlags)
+        c.acorr = ax.acorr(data, detrend=detrend, usevlines=True, 
+	  maxlags=maxlags, normed=True)
+	ax.set_ylabel('acorr')
 
         ax = c.ax4 = fig.add_subplot(N,1,4)
         c.psd = ax.psd(data, Fs=Fs, detrend=detrend)
+	ax.set_ylabel('psd')
 
         ax = c.ax5 = fig.add_subplot(N,1,5)
         c.specgtram = ax.specgram(data, Fs=Fs, detrend=detrend)
+	ax.set_ylabel('specgram')
         return c
 
 
@@ -111,6 +117,9 @@
 
     desc = Descriptives(data)
     print desc
-    c = desc.plots(pylab.figure, Fs=12, fmt='-o')
+    c = desc.plots(pylab.figure, Fs=12, fmt='-')
     c.ax1.set_title(fname)
+
+    c.fig.savefig('stats_descriptives.png', dpi=150)    
+    c.fig.savefig('stats_descriptives.eps')    
     pylab.show()

Modified: trunk/py4science/examples/stats_distributions.py
===================================================================
--- trunk/py4science/examples/stats_distributions.py	2007-10-26 20:15:42 UTC (rev 4028)
+++ trunk/py4science/examples/stats_distributions.py	2007-10-26 20:15:51 UTC (rev 4029)
@@ -35,18 +35,18 @@
 # 1/lambda.  Plot all three on the same graph and make a legend.
 # Decorate your graphs with an xlabel, ylabel and title
 fig = figure()
-ax = fig.add_subplot(111)
+ax = fig.add_subplot(311)
 p, bins, patches = ax.hist(wait_times, 100, normed=True)
 l1, = ax.plot(bins, rate*numpy.exp(-rate * bins), lw=2, color='red')
 l2, = ax.plot(bins, scipy.stats.expon.pdf(bins, 0, 1./rate),
         lw=2, ls='--', color='green')
-ax.set_xlabel('waiting time')
+
 ax.set_ylabel('PDF')
-ax.set_title('waiting time density of a %dHz Poisson emitter'%rate)
+ax.set_title('waiting time densities of a %dHz Poisson emitter'%rate)
+ax.text(0.05, 0.9, 'one interval', transform=ax.transAxes)
 ax.legend((patches[0], l1, l2), ('simulated', 'analytic', 'scipy.stats.expon'))
 
 
-
 # plot the distribution of waiting times for two events; the
 # distribution of waiting times for N events should equal a N-th order
 # gamma distribution (the exponential distribution is a 1st order
@@ -54,17 +54,15 @@
 # Hint: you can stride your emission times array to get every 2nd
 # emission
 wait_times2 = numpy.diff(emit_times[::2])
-fig = figure()
-ax = fig.add_subplot(111)
+ax = fig.add_subplot(312)
 p, bins, patches = ax.hist(wait_times2, 100, normed=True)
 l1, = ax.plot(bins, scipy.stats.gamma.pdf(bins, 2, 0, 1./rate),
         lw=2, ls='-', color='red')
-ax.set_xlabel('2 event waiting time 2 events')
+
 ax.set_ylabel('PDF')
-ax.set_title('waiting time density of a %dHz Poisson emitter'%rate)
+ax.text(0.05, 0.9, 'two intervals', transform=ax.transAxes)
 ax.legend((patches[0], l1), ('simulated', 'scipy.stats.gamma'))
 
-
 # plot the distribution of waiting times for 10 events; again the
 # distribution will be a 10th order gamma distribution so plot that
 # along with the empirical density.  The central limit thm says that
@@ -81,19 +79,20 @@
 mu, var = 10*expon_mean, 10*expon_var
 sigma = numpy.sqrt(var)
 wait_times10 = numpy.diff(emit_times[::10])
-fig = figure()
-ax = fig.add_subplot(111)
+ax = fig.add_subplot(313)
 p, bins, patches = ax.hist(wait_times10, 100, normed=True)
 l1, = ax.plot(bins, scipy.stats.gamma.pdf(bins, 10, 0, 1./rate),
         lw=2, ls='-', color='red')
 l2, = ax.plot(bins, scipy.stats.norm.pdf(bins, mu, sigma),
         lw=2, ls='--', color='green')
 
-ax.set_xlabel('waiting time 10 events')
+ax.set_xlabel('waiting times')
 ax.set_ylabel('PDF')
-ax.set_title('10 event waiting time density of a %dHz Poisson emitter'%rate)
+ax.text(0.1, 0.9, 'ten intervals', transform=ax.transAxes)
 ax.legend((patches[0], l1, l2), ('simulated', 'scipy.stats.gamma', 'normal approx'))
 
+fig.savefig('stats_distributions.png', dpi=150)
+fig.savefig('stats_distributions.eps')
 
 
 show()


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.