In [47]:

    

%matplotlib inline
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
plt.style.use('bmh')
pd.set_option('display.max_rows', 200)


    

In [51]:

    

df = pd.read_csv('timing.txt', sep='\t')
#print(df)


    

In [55]:

    

fig = plt.figure(figsize=(16,12))
for ncomm in df['Ncomm'].unique():
    for ncells in df['Ncells'].unique():
        for npart in df['Nparticles'].unique():
        #for npart in [99856]:
            mask = np.logical_and(df['Nparticles']==npart, np.logical_and(df['Ncomm']==ncomm, df['Ncells']==ncells))
            x = df[mask]['Nranks']
            y = df[mask]['Time']
            plt.plot(x, y, label='%s %s %s' % (ncomm, ncells, npart))
plt.legend()
plt.yscale('log')
plt.xscale('log')
plt.xlim(1,12)


    

    
Out[55]:





(1, 12)






    

In [13]:

    

for n in df['Ncomm'].unique():
    x = df[df['Ncomm']==n]['Nranks']
    y = df[df['Ncomm']==n]['Time']
    plt.plot(x, y, label=n)
plt.legend()


    

    
Out[13]:






  

    

      

      
Nparticles
      
Ncells
      
Nranks
      
Time
    
  
  

    

      
0
      
19881
      
210
      
1
      
6.70221
    
    

      
1
      
19881
      
210
      
2
      
4.26774
    
    

      
2
      
19881
      
210
      
3
      
3.49359
    
    

      
3
      
19881
      
210
      
4
      
3.03752
    
    

      
4
      
19881
      
210
      
5
      
2.79173
    
    

      
5
      
19881
      
210
      
6
      
2.74210
    
    

      
6
      
19881
      
210
      
7
      
2.79696
    
    

      
7
      
19881
      
210
      
8
      
2.95650
    
    

      
8
      
19881
      
210
      
9
      
3.17641
    
    

      
9
      
19881
      
210
      
10
      
3.22525
    
    

      
10
      
19881
      
210
      
11
      
3.46706
    
    

      
11
      
19881
      
210
      
12
      
3.50533
    
  

In [18]:

    

dt = 0.006
Ndt = int(1/dt)
tmax = 5
1 + int(tmax / dt / Ndt)


    

    
Out[18]:





6

In [ ]: