Cjl

bin/plotSingleClusterBindingCurve.py

@@ -35,41 +35,65 @@
 
 #set up command line argument parser
 parser = argparse.ArgumentParser(description='bootstrap fits')
-parser.add_argument('-b', '--binding_curves', required=True, metavar=".CPseries.pkl",
+parser.add_argument('-b', '--binding_curves', required=True, metavar='.CPseries.pkl',
                    help='file containining the binding curve information')
-parser.add_argument('-f', '--single_cluster_fits', required=True, metavar=".CPfitted.pkl",
+parser.add_argument('-f', '--single_cluster_fits', required=True, metavar='CPfitted.pkl',
                    help='file containining the single cluster fits')
-parser.add_argument('-c', '--concentrations', required=True, metavar="concentrations.txt",
+parser.add_argument('-c', '--concentrations', required=True, metavar='concentrations.txt',
                     help='text file giving the associated concentrations')
-parser.add_argument('-i', '--cluster', required=True, metavar="clusterID",
-                    help='cluster Id for which to plot')
-
-parser.add_argument('-out', '--out_file', 
-                   help='output filename. default is "cluster_X.binding_curve.pdf"')
-
+parser.add_argument('-i', '--cluster', required=False, metavar='clusterID', action='append',
+                    help='individual clusterIDs to plot')
+parser.add_argument('-if', '--cluster_file', required=False, metavar='clusterID file', help='text file containing clusterIDs to plot, one-per-line')
+parser.add_argument('-out', '--out_dir', default='plots',
+                   help='output directory. default is "plots"')
 
 
 if __name__ == '__main__':
     args = parser.parse_args()
 
+    if not len(sys.argv) > 1: # Print help if this script is called with no arguments (equivalent to -h)
+        parser.print_help()
+
     bindingCurveFilename  = args.binding_curves
     singleClusterFilename = args.single_cluster_fits
-    cluster = args.cluster
-
+
+    # Read in cluster file
+    clusters = []
+    if args.cluster_file is not None:
+        with open(args.cluster_file) as f:
+            clusters = f.readlines()
+            clusters = [line.strip() for line in clusters] # Strip trailing newline characters and whitespace
+
+    # Add any individual clusterIDs that were passed in with -i
+    if args.cluster is not None:
+        clusters += args.cluster
+
+    if len(clusters) <= 0:
+        sys.exit('ERROR: No clusterIDs were passed in. Use -i or -if to indicate which clusterIDs should be plotted. Exiting...')
+
     # load data
     concentrations = np.loadtxt(args.concentrations)
     bindingSeries =  pd.read_pickle(bindingCurveFilename)
     fittedSingles = pd.read_pickle(singleClusterFilename)
 
-    # plot
-    fig = plt.figure(figsize=(4,3))
-    ax = fig.add_subplot(111)
-    fitParameters = pd.DataFrame(columns=['fmax', 'dG', 'fmin'])
-    fitFun.plotFitCurve(concentrations,
-                        bindingSeries.loc[cluster],
-                        fittedSingles.loc[cluster],
-                        fitParameters, ax=ax)
+    # Prepare the output directory
+    try:
+        os.makedirs(args.out_dir)
+    except OSError:
+        if not os.path.isdir(args.out_dir):
+            raise
 
-    if args.out_file is None:
-        args.out_file = 'cluster_%s.binding_curve.pdf'%cluster
-    plt.savefig( args.out_file)
+    # Iterate through each ClusterID    
+    for currClusterID in clusters:
+
+        # plot
+        fig = plt.figure(figsize=(4,3))
+        ax = fig.add_subplot(111)
+        fitParameters = pd.DataFrame(columns=['fmax', 'dG', 'fmin'])
+        plotFun.plotFitCurve(concentrations,
+                            bindingSeries.loc[currClusterID],
+                            fittedSingles.loc[currClusterID],
+                            fitParameters, ax=ax)
+
+        currOutputFilename = os.path.join(args.out_dir, 'cluster_%s.binding_curve.pdf'%currClusterID)
+        plt.savefig(currOutputFilename)