#%install_ext https://raw.github.com/minrk/ipython_extensions/master/extensions/nbtoc.py
#%load_ext nbtoc
%autosave 60

#%nbtoc

import pandas as pd
import matplotlib.pyplot

# somehow pandas has trouble reading the sorted frequency table
a = pd.read_table("sorted_freq.txt", names = ['col']) 
c = []
for i in range(a.shape[0]):
    b = a['col'][i].split()
    if (len(b) == 2):
        if (b[1] != 'NA') and (b[1]!='ArrDelay'):
            c.append([float(b[0]), float(b[1])])
c = np.array(c).transpose()

delay = c[1]
freq = c[0]

plt.plot(delay, freq, '.b')
plt.xlabel('Delay time (min)', size=14)
plt.ylabel('Frequency', size=14)
xlim_low = -100
xlim_high = 100
plt.title('Truncated view of frequency table from'+\
          '{0} to {1}'.format(xlim_low, xlim_high), size=15)

plt.xlim(xlim_low, xlim_high)
plt.axvline('6.566504', label = 'mean', color = 'r', ls = ':', lw = 3 )
plt.axvline('0.00', label = 'median', color = 'g', ls = '--', lw = 3 )
plt.legend(loc = 'best')

plt.plot(delay, freq, '.b')
plt.xlabel('Delay time (min)', size=14)
plt.ylabel('Frequency', size=14)
xlim_low = -100
xlim_high = 100
plt.title('Total view of frequency table', size=15)
plt.axvline('6.566504', label = 'mean', color = 'r', ls = ':', lw = 3)
plt.axvline('0.00', label = 'median', color = 'g', ls = '--', lw = 3 )
plt.legend(loc = 'best')

print 'each line should at most take '+\
'{0}s'.format( 30 * 60 / (1e6))

y = [1, 2, 3]
result1 = [6.566504, 0.00000, 31.556326]
result2 = [6.56650421703, 0.0, 31.5563262623]
result3 = [6.591345,  0.000000, 31.465116]

plt.plot(result1, y, 'rs', markeredgewidth=1, 
         alpha = .25, markersize = 10, label="Method 1")
plt.plot(result2, y, 'gx', markeredgewidth=3, 
         alpha = .25, markersize = 10, label="Method 2")
plt.plot(result3, y, 'b*', markeredgewidth=1, 
         alpha = .25, markersize = 10, label="Method 3")

plt.legend(loc='lower right')
plt.title('Comparison of statistics from different approaches', size = 15)
label = ['mean', 'median', 'std. dev',]
plt.yticks( y, label, size = 15)
plt.margins(0.2)
plt.xlabel('Arrival delay (min)', size = 15)

plt.title('Comparison of runtime of different methods', size = 16)

y = [1, 2, 3]
time_taken = [282.900/60., 181.094/60., 269.420/60.] 

plt.plot(time_taken, x, 'o')
label = ['Freq table - Shell + R ', 'Python Pandas', 'FastCSVSample + R',]
plt.yticks( y, label, size = 12)
plt.margins(0.2)
plt.xlim(0,6)
plt.xlabel('Wall clock time (min)', size = 15)

!cat ../freq_count.sh

!cat ../method1.R

!cat ../compute_stat.py

!cat ../method2.R

!cat ../method3.R

!cat ../NotSoFastCSVSample/R/csvSample.R