Add functionality to histogram functions in core

icaro/core/hst_functions.py

@@ -0,0 +1,124 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Created on Fri Nov  3 20:21:59 2017
+
+Decorate histogram function in IC hist_functions module
+
+add statitics into the plots
+define canvas to subdivide figure in subplots
+
+@author: hernando
+"""
+
+import numpy             as np
+import scipy.stats       as stats
+import matplotlib.pyplot as plt
+import invisible_cities.icaro.hst_functions as hst
+from invisible_cities.icaro.hst_functions import shift_to_bin_centers
+from invisible_cities.icaro.hst_functions import resolution, gausstext
+from invisible_cities.icaro.hst_functions import labels
+from invisible_cities.icaro.hst_functions import save_to_folder
+
+class Canvas:
+    """ class object to store subplot in figure
+    """
+    def __init__(self, nx, ny, nxsize=5.4, nysize=6.2):
+        """ divide the figure in (nx, ny) subplots
+        inputs:
+            nx, ny : int, int
+                     number of subplots in (x, y)
+            nxsize, nysize: float, flot
+                    x, y size of the subplots
+        """
+        self.nx = nx
+        self.ny = ny
+        self.n  = 1
+        plt.figure(figsize=(nysize*ny, nxsize*nx))
+        plt.subplot(nx, ny, self.n)
+
+    def __call__(self, i):
+        """ locate in subplot i in figure
+        """
+        plt.subplot(self.nx, self.ny, i)
+        return True
+
+    def __add__(self, i):
+        """ locate to next i plot in figure
+        """
+        self.n += i
+        plt.subplot(self.nx, self.ny, n)
+        return True
+
+
+def _xypos(xs, ys, xf=0.1, yf=0.7):
+    x0, dx = min(xs), max(xs)-min(xs)
+    y0, dy = min(ys), max(ys)-min(ys)
+    xp = x0 + xf*dx
+    yp = y0 + yf*dy
+    return (xp, yp)
+
+def _hist(xs, bins=100, range=None, stats=('entries', 'mean', 'rms'),
+         xylabels = (), stats_xypos=(0.1, 0.7),
+         *args, **kargs):
+    """
+    hist function extended to plot stattistics
+    inputs:
+        stats: (string,) or None. Default: ('entries', 'mean', 'rms')
+            name of the statistics parameters to be into the histogram
+            if none use None
+            possibilities: 'total entries'
+        stats_xypos: (float, float) (default (0.1, 0.7))
+            (x, y) relative positions in the plot,
+            x, y values in rage (0., 1.)
+    outputs:
+        none
+    """
+    if (range==None):
+        range = (np.min(xs), np.max(xs))
+    cc = hst.hist(xs, bins=bins, range=range, *args, **kargs);
+    if (not stats):
+        return cc
+    ys, xedges = np.histogram(xs, bins, range=range)
+    ns = len(xs)
+    sel = np.logical_and(xs >= range[0], xs <= range[1])
+    nos, mean, rms = len(xs[sel]), np.mean(xs[sel]), np.std(xs[sel])
+    epsilon = (1.*nos)/(1.*ns)
+    ss = ''
+    if ('total entries') in stats:
+        ss += 'total entries  {0:d} \n'.format(ns)
+    if ('entries') in stats:
+        ss += 'entries {0:d} \n'.format(nos)
+    if ('mean') in stats:
+        ss += 'mean {0:.3f} \n'.format(mean)
+    if ('rms') in stats:
+        ss += 'rms  {0:.3f} \n'.format(rms)
+    xp, yp = _xypos(xedges, ys, xf=stats_xypos[0], yf=stats_xypos[1])
+    plt.text(xp, yp, ss)
+    return cc
+
+
+def _decorate(fun):
+    def _decorator(*args, **kargs):
+        if ('canvas' in kargs):
+            kargs['new_figure'] = False
+            del kargs['canvas']
+        has_xylabels, xylabels = 'xylabels' in kargs, ()
+        if (has_xylabels):
+            xylabels = kargs['xylabels']
+            del kargs['xylabels']
+        response = fun(*args, **kargs)
+        if (len(xylabels)>0):
+            hst.labels(*xylabels)
+        return response
+    return _decorator
+
+hist                = _decorate(_hist)
+plot                = _decorate(hst.plot)
+hist2d              = _decorate(hst.hist2d)
+errorbar            = _decorate(hst.errorbar)
+pdf                 = _decorate(hst.pdf)
+scatter             = _decorate(hst.scatter)
+profile_and_scatter = _decorate(hst.profile_and_scatter)
+hist2d_profile      = _decorate(hst.hist2d_profile)
+display_matrix      = _decorate(hst.display_matrix)