Added a relatively "deep" example

astrodendro/plot.py

@@ -190,7 +190,7 @@ def get_lines(self, structures=None, subtree=True, **kwargs):
         # Case 2: one structure is selected, and subtree is True
         else:
             if subtree:
-                if type(structures[0]) is int:
+                if isinstance(structures[0], int):
                     structure = self.dendrogram[structures[0]]
                 else: 
                     structure = structures[0]

docs/examples.rst

@@ -0,0 +1,88 @@
+Example Uses of astrodendro
+===========================
+
+All of these examples use the Perseus data set
+
+Example 1
+---------
+
+Extract the dendrogram structure using a minimum-peak cutoff, then make some
+intricate plots.
+
+These plots show two derived quantities - the "flux" and the effective radius -
+plotted against each other in two different ways.  In order to visualize the
+structure as plotted, the lines are color- and marker- coded to match each
+other.
+
+
+.. plot::
+   :include-source:
+
+    from astrodendro import Dendrogram
+    from astropy.io import fits
+    from astrodendro.pruning import min_peak
+    from astrodendro.analysis import PPStatistic
+    import itertools
+    import pylab as pl
+    import numpy as np
+
+    image = fits.getdata('PerA_Extn2MASS_F_Gal.fits')
+
+    from astropy import units as u
+    # this metadata isn't really appropriate for this data set, but it is
+    # useful for the example
+    metadata = {}
+    metadata['data_unit'] = u.Jy / u.beam
+    metadata['spatial_scale'] =  6 * u.arcsec
+    metadata['beam_major'] =  22.9 * u.arcsec # FWHM
+    metadata['beam_minor'] =  22.9 * u.arcsec # FWHM
+
+    d = Dendrogram.compute(image, min_value=1.0, min_delta=1., min_npix=10,
+                           is_independent=min_peak(3), verbose=True)
+
+    f1 = pl.figure(1,figsize=(15,5))
+    f1.clear()
+    ax1,ax2,ax3 = (pl.subplot(1,3,ii) for ii in xrange(1,4))
+
+    p = d.plotter()
+    p.plot_tree(ax3, color='k', alpha=0.5)
+
+    # quantities to plot:
+    flux = {}
+    radius = {}
+
+    markers = itertools.cycle(['o','h','s','^','v','<','>','p','D','*'])
+
+    for leaf in d.leaves:
+        id = leaf.idx
+        print("Plotting ID ",id)
+        # need one suite of plotted things for each leaf
+        flux[id] = []
+        radius[id] = []
+
+        while hasattr(leaf,'parent'):
+            stat = PPStatistic(leaf, metadata=metadata)
+            flux[id].append(stat.flux.value)
+            radius[id].append(stat.radius.value)
+            leaf = leaf.parent
+
+        L, = ax1.loglog(np.array(radius[id])**2 * np.pi,
+                   flux[id],
+                   '-',marker=markers.next(),
+                   alpha=0.5)
+        ax2.loglog(np.array(radius[id])**2 * np.pi,
+                   np.array(flux[id]) / (4/3. * np.pi * np.array(radius[id])**2),
+                   'o-', color=L.get_color(), marker=L.get_marker(),
+                   alpha=0.5)
+        p.plot_tree(ax3, structure=id, color=L.get_color(), linewidth=2, alpha=0.5)
+        lines = p.get_lines(structure=id, color=L.get_color(), linewidth=2, alpha=0.5)
+        ax3.plot(*lines.get_segments()[0].T,marker=L.get_marker())
+
+
+
+    ax1.set_xlabel('Area (arcsec$^2$)')
+    ax1.set_ylabel('Flux (Jy)')
+    ax2.set_xlabel('Area (arcsec$^2$)')
+    ax2.set_ylabel('"Column Density" (Jy/arcsec$^3$)')
+    ax3.set_xlabel("Structure")
+    ax3.set_ylabel("Flux")

docs/index.rst

@@ -36,6 +36,8 @@ Mapping these terms back onto the structure gives the following:
 For an example of use of dendrograms on real data, see `Goodman, A. et al
 (2009) <http://adsabs.harvard.edu/abs/2009Natur.457...63G>`_.
 
+For some examples of dendrograms in practice, see :doc:`examples`
+
 Documentation
 -------------