diff --git a/pyglance/glance/data.py b/pyglance/glance/data.py
index 0a5d36bf56099f4729ecabaa663a889fd484d7a0..9f9c71b1f0052f53af99280ed64e4f4a1adb4929 100644
--- a/pyglance/glance/data.py
+++ b/pyglance/glance/data.py
@@ -8,6 +8,7 @@ Copyright (c) 2010 University of Wisconsin SSEC. All rights reserved.
"""
import logging
+import os, subprocess, datetime
import numpy as np
import glance.delta as delta
@@ -277,6 +278,78 @@ class DiffInfoObject (object) :
return diff_data_object
+class FileInfo (object) :
+ """
+ This class represents information about a file object. It may or may not include the actual file object.
+
+ The following member variables are available from this class:
+
+ path - the file path to reach the original file on disk
+ md5_sum - an md5 sum calculated from the original file
+ last_modified - the time that the file was last modified (TODO, what form should this be in?)
+ file_object - the file object that can be used to access the data in the file, may be None
+ """
+
+ def __init__(self, pathToFile, md5sum=None, lastModifiedTime=None, fileObject=None) :
+ """
+ Create the file info object using the values given.
+
+ If the md5 sum and last modified time aren't given, the initialization will figure them out.
+ Note: if the md5 sum is not given, the file object will also be loaded.
+ """
+
+ self.path = pathToFile
+
+ # if the file doesn't exist, stop
+ # TODO, is this the right strategy?
+ if not os.path.exists(self.path) :
+ LOG.warn("Requested file " + self.path + " could not be opened because it does not exist.")
+ self.md5_sum = None
+ self.last_modified = None
+ self.file_object = None
+ return
+
+ # if the md5 sum isn't given, load the file and figure it out
+ if md5sum is None:
+
+ # open the file
+ LOG.info("Opening " + self.path)
+ tempPath = os.path.abspath(os.path.expanduser(self.path))
+ LOG.debug("Provided path after normalization and symbol expansion: " + tempPath)
+ fileObject = io.open(tempPath, allowWrite=allowWrite)
+
+ # figure out the md5 sum
+ tempSubProcess = subprocess.Popen("md5sum \'" + tempPath + "\'", shell=True, stdout=subprocess.PIPE)
+ md5sum = tempSubProcess.communicate()[0].split()[0]
+ LOG.info("File md5sum: " + str(md5sum))
+
+ self.md5_sum = md5sum
+ self.file_object = fileObject
+
+ # if the last modified time isn't given, figure it out
+ if lastModifiedTime is None :
+
+ statsForFile = os.stat(os.path.abspath(os.path.expanduser(self.path)))
+ lastModifiedTime = datetime.datetime.fromtimestamp(statsForFile.st_mtime).ctime() # should time zone be forced?
+ LOG.info ("File was last modified: " + lastModifiedTime)
+
+ self.last_modified = lastModifiedTime
+
+ def get_version_without_file_object (self) :
+ """
+ get a version of this object without a file object
+ (this method is useful if you want file information but do not need access and want to save space)
+ """
+ toReturn = None
+
+ if self.file_object is None:
+ toReturn = self
+ else:
+ toReturn = FileInfo(self.path, self.md5_sum, self.last_modified)
+
+ return toReturn
+
+
if __name__=='__main__':
import doctest
doctest.testmod()
diff --git a/pyglance/glance/figures.py b/pyglance/glance/figures.py
index ef5035d97e42c919f711c5fb5897444d4b3a5af1..05d561584fd8070032ef0dd692ca228d59edeb59 100644
--- a/pyglance/glance/figures.py
+++ b/pyglance/glance/figures.py
@@ -13,13 +13,14 @@ matplotlib.use('Agg') # use the Anti-Grain Geometry rendering engine
from pylab import *
+import matplotlib.cm as cm
import matplotlib.pyplot as plt
import matplotlib.colors as colors
-from matplotlib.ticker import FormatStrFormatter
+from matplotlib.ticker import FormatStrFormatter
import logging
import numpy as np
-from numpy import ma
+from numpy import ma
import glance.graphics as maps
import glance.delta as delta
@@ -236,6 +237,34 @@ def create_scatter_plot(dataX, dataY, title, xLabel, yLabel, badMask=None, epsil
return figure
+# build a hexbin plot of the x,y points and show the density of the point distribution
+def create_hexbin_plot(dataX, dataY, title, xLabel, yLabel) :
+
+ # make the figure
+ figure = plt.figure()
+ axes = figure.add_subplot(111)
+
+ # the hexbin plot of the good data
+ plt.hexbin(dataX, dataY, bins='log', cmap=cm.jet)
+ plt.axis([dataX.min(), dataX.max(), dataY.min(), dataY.max()])
+
+ # create a color bar
+ cb = plt.colorbar()
+ cb.set_label('log10 (count + 1)')
+
+ # and some informational stuff
+ axes.set_title(title)
+ plt.xlabel(xLabel)
+ plt.ylabel(yLabel)
+
+ # format our axes so they display gracefully
+ yFormatter = FormatStrFormatter("%4.4g")
+ axes.yaxis.set_major_formatter(yFormatter)
+ xFormatter = FormatStrFormatter("%4.4g")
+ axes.xaxis.set_major_formatter(xFormatter)
+
+ return figure
+
# build a histogram figure of the given data with the given title and number of bins
def create_histogram(data, bins, title, xLabel, yLabel, displayStats=False) :
diff --git a/pyglance/glance/plotcreatefns.py b/pyglance/glance/plotcreatefns.py
index b5505ff102337843d013dbf7f77844e9a2841fd6..d43bc03f74fabb6a7be319ac677755a0bff302b5 100644
--- a/pyglance/glance/plotcreatefns.py
+++ b/pyglance/glance/plotcreatefns.py
@@ -273,6 +273,18 @@ class BasicComparisonPlotsFunctionFactory (PlottingFunctionFactory) :
"scatter plot of file a values vs file b values for " + variableDisplayName,
"Scatter.png", compared_fig_list)
+ # make a hexplot, which is like a scatter plot with density
+ if ('do_plot_hex' not in doPlotSettingsDict) or (doPlotSettingsDict['do_plot_hex']) :
+
+ assert(aData.shape == bData.shape)
+ assert(bData.shape == goodInBothMask.shape)
+
+ functionsToReturn['scatterD'] = ((lambda : figures.create_hexbin_plot(aData[goodInBothMask], bData[goodInBothMask],
+ "Value in File A vs Value in File B",
+ "File A Value", "File B Value")),
+ "density of file a values vs file b values for " + variableDisplayName,
+ "Hex.png", compared_fig_list)
+
return functionsToReturn
"""