Plotting now works in general for all datasets. plot_throughput only …

…has to be run in the current directory where the data resides.

plot_throughput.py

@@ -6,11 +6,13 @@
 @author: Damian
 """
 import numpy as np
-import cPickle
+#import cPickle
 import matplotlib
 matplotlib.use("Agg")
 import matplotlib.lines as lines
 import matplotlib.pyplot as plt
+import glob
+import re
 from subprocess import call
 import h5py
 
@@ -19,72 +21,67 @@
 ROADLENGTH = 50
 TRIALS = 50
 
-area = 1*(REAL_LANES)*ROADLENGTH
-speed = 0
-size = -1
-last = -1
+AREA = 1*(REAL_LANES)*ROADLENGTH
+SPEED = 0
+SIZE = -1
+LAST = -1
 
-def throughputnorm(trial, sizes):
-    x= [vehicles[speed]/ROADLENGTH for vehicles in trial[-1]]
-    return 1.*np.sum(x)/len(x)
+def throughput(_trial):
+    '''Computes the throughput of one trial of the simulation.'''
+    return np.sum([vehicles[SPEED]/ROADLENGTH for vehicles in _trial[LAST]])
 
-def throughput(trial, sizes):
-    return np.sum([vehicles[speed]/ROADLENGTH for vehicles, s in zip(trial[last],
-                   sizes[last]) if s])
-def through(vehicles):
-    return 1.*np.sum(i[0] for i in vehicles)/ROADLENGTH/len(vehicles)
-
-def load(ratio, density):
-    throughput = np.array([], dtype=np.int8)
-    filename = "CarRatio.%.2f_Density.%.2f.h5" % (ratio,density)
+def load(_ratio, _density):
+    '''Loads data from the hdf5 dataset.'''
+    vehicledata = np.array([], dtype=np.int8)
+    filename = "CarRatio.%.2f_Density.%.2f.h5" % (_ratio, _density)
     call(['bunzip2', filename+'.bz2'])
     fid = h5py.File(filename, 'r')
-    for t in xrange(TRIALS):
-        group = "CarRatio::%.2f/Density::%.2f/" % (ratio, density)
-        trial = "Trial::%04d" % (t+1)
-        dset = fid[group+trial]
-        throughput = np.append(throughput, dset)
-
-    throughput = np.reshape(throughput, (TRIALS, dset.shape[0],
+    for n in xrange(TRIALS):
+        group = "CarRatio::%.2f/Density::%.2f/" % (_ratio, _density)
+        _trial = "Trial::%04d" % (n+1)
+        dset = fid[group+_trial]
+        vehicledata = np.append(vehicledata, dset)
+    vehicledata = np.reshape(vehicledata, (TRIALS, dset.shape[0],
                                          dset.shape[1],
                                          dset.shape[2]))
     fid.close()
     call(["bzip2", "-6", filename])
-    return throughput
-
-
+    return vehicledata
 
-fluxes = np.zeros([1,19])
-variance = np.zeros([1,19, 2])
-for x, ratio in enumerate(np.arange(1)):#0.1,1,0.1)):
-    if ratio == 1:
-        ratio = 1
-    if ratio == 0:
-        ratio = 0
-    for y, density in enumerate(np.arange(0.05,1,0.05)):
-        vehicledata = load(ratio, density)
-        sizedata = vehicledata[:,2001:,:,size]
-        vehicledata = np.cumsum(vehicledata[:,2001:,:,:], axis=1)
-        vehicleflux = [throughput(trial,sizes)
-                        for trial, sizes in zip(vehicledata, sizedata)]
-        fluxes[x,y] = np.mean(vehicleflux)
-        variance[x,y] = [np.max(vehicleflux), np.min(vehicleflux)]
+if __name__ == "__main__":
+    FILES = glob.glob("*.bz2")
+    DENSITIES = np.arange(0.05, 1, 0.05)
+    RATIOS = np.array([])
+    for i in FILES:
+        car_ratio = float(re.findall(r"CarRatio.(\d.\d+)", i)[0])
+        if car_ratio not in RATIOS:
+            RATIOS = np.append(RATIOS, car_ratio)
 
+    THROUGHPUT = np.zeros([len(RATIOS), len(DENSITIES)])
+    STDEV = np.zeros([len(RATIOS), len(DENSITIES)])
+    for x, ratio in enumerate(np.arange(1)):#0.1,1,0.1)):
+        for y, density in enumerate(DENSITIES):
+            data = load(ratio, density)
+            # sizedata = vehicledata[:,2001:,:,SIZE]
+            data = np.cumsum(data[:, :, :, :], axis=1)
+            flux = [throughput(trial)
+                            for trial in data]
+            THROUGHPUT[x, y] = np.mean(flux)
+            STDEV[x, y] = np.std(flux)
 
-markers = list(lines.Line2D.markers.keys())
-markers.sort()
-N = len(markers)
-labels = np.arange(0.1, 1, 0.1)
-for ydata, label, i in zip(fluxes, labels, range(len(labels))):
-    marker = markers[i%N]
-#    plt.plot(np.arange(0.05,1,0.05), ydata, label=label, marker=marker)
-    plt.errorbar(np.arange(0.05,1,0.05), ydata, yerr=np.abs(variance[i].T-ydata),label=label, marker=marker)
-#plt.legend(loc="upper left", bbox_to_anchor=(1,1))
-#plt.legend()
-plt.xlabel('road density', size=18)
-plt.ylabel('number of exiting vehicles', size=18)
-#plt.ylim(0,2000)
-#plt.xlim(0,1.2)
-plt.title(r'Two Real lanes with adaptation', size=18)
+    __markers__ = list(lines.Line2D.markers.keys())
+    __markers__.sort()
+    N = len(__markers__)
+    __labels__ = np.arange(0.1, 1, 0.1)
+    for ydata, label, i in zip(THROUGHPUT, __labels__, range(len(__labels__))):
+        marker = __markers__[i % N]
+        plt.errorbar(DENSITIES, ydata, 
+                    yerr=STDEV[i, :], label=label, marker=marker)
+    plt.legend(loc="upper left", bbox_to_anchor=(1, 1))
+    plt.xlabel('road density', size=18)
+    plt.ylabel('number of exiting vehicles', size=18)
+    #plt.ylim(0,2000)
+    #plt.xlim(0,1.2)
+    plt.title(r'Two Real lanes with adaptation', size=18)
 
-plt.savefig("throughput.png")
+    plt.savefig("throughput.png")