In [1]:

    

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline


    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [2]:

    

# Import stats
FILE = 'cleaned_stats.csv'
# JSON = '../stats.json'
JSON = '../stats50.json'


    

In [3]:

    

#df = pd.read_csv(FILE)
df = pd.read_json(JSON)


    

In [4]:

    

del df['board']
df.head()


    

    
Out[4]:






  

    

      

      
dimension
      
heuristic
      
rand_degree
      
steps
      
time
    
  
  

    

      
0
      
5
      
1
      
15
      
NaN
      
1067401.680
    
    

      
1
      
5
      
2
      
15
      
2164.0
      
44692.390
    
    

      
2
      
5
      
1
      
15
      
NaN
      
1043000.295
    
    

      
3
      
5
      
2
      
15
      
NaN
      
1059986.975
    
    

      
4
      
5
      
1
      
15
      
NaN
      
1026634.820
    
  

In [5]:

    

df['time_m'] = df['time']/60000
df.head()


    

    
Out[5]:






  

    

      

      
dimension
      
heuristic
      
rand_degree
      
steps
      
time
      
time_m
    
  
  

    

      
0
      
5
      
1
      
15
      
NaN
      
1067401.680
      
17.790028
    
    

      
1
      
5
      
2
      
15
      
2164.0
      
44692.390
      
0.744873
    
    

      
2
      
5
      
1
      
15
      
NaN
      
1043000.295
      
17.383338
    
    

      
3
      
5
      
2
      
15
      
NaN
      
1059986.975
      
17.666450
    
    

      
4
      
5
      
1
      
15
      
NaN
      
1026634.820
      
17.110580
    
  

In [6]:

    

df.columns = ['Dimension','Heuristic', 'Rand_degree', 'Steps', 'Time in ms', 'Time in minutes']


    

In [7]:

    

df.groupby('Heuristic').describe(include=[np.number])


    

    
Out[7]:






  

    

      

      

      
Dimension
      
Rand_degree
      
Steps
      
Time in minutes
      
Time in ms
    
    

      
Heuristic
      

      

      

      

      

      

    
  
  

    

      
1
      
count
      
900.000000
      
900.000000
      
737.000000
      
900.000000
      
9.000000e+02
    
    

      
mean
      
4.000000
      
10.000000
      
1142.120760
      
2.634366
      
1.580620e+05
    
    

      
std
      
0.816951
      
4.084753
      
2202.493057
      
5.685925
      
3.411555e+05
    
    

      
min
      
3.000000
      
5.000000
      
2.000000
      
0.000012
      
6.950000e-01
    
    

      
25%
      
3.000000
      
5.000000
      
6.000000
      
0.000053
      
3.182500e+00
    
    

      
50%
      
4.000000
      
10.000000
      
17.000000
      
0.004524
      
2.714175e+02
    
    

      
75%
      
5.000000
      
15.000000
      
921.000000
      
1.394405
      
8.366430e+04
    
    

      
max
      
5.000000
      
15.000000
      
9737.000000
      
37.801713
      
2.268103e+06
    
    

      
2
      
count
      
900.000000
      
900.000000
      
787.000000
      
900.000000
      
9.000000e+02
    
    

      
mean
      
4.000000
      
10.000000
      
783.714104
      
1.811633
      
1.086980e+05
    
    

      
std
      
0.816951
      
4.084753
      
1878.861366
      
4.641229
      
2.784737e+05
    
    

      
min
      
3.000000
      
5.000000
      
2.000000
      
0.000011
      
6.550000e-01
    
    

      
25%
      
3.000000
      
5.000000
      
5.000000
      
0.000038
      
2.265000e+00
    
    

      
50%
      
4.000000
      
10.000000
      
10.000000
      
0.000119
      
7.137500e+00
    
    

      
75%
      
5.000000
      
15.000000
      
329.000000
      
0.131653
      
7.899185e+03
    
    

      
max
      
5.000000
      
15.000000
      
9995.000000
      
31.918191
      
1.915091e+06
    
  

In [6]:

    

dn = df[np.isfinite(df['steps'])]
dn.groupby('heuristic').describe(include=[np.number])


    

    
Out[6]:






  

    

      

      

      
dimension
      
rand_degree
      
steps
      
time
    
    

      
heuristic
      

      

      

      

      

    
  
  

    

      
1
      
count
      
737.000000
      
737.000000
      
737.000000
      
737.000000
    
    

      
mean
      
3.905020
      
8.907734
      
1142.120760
      
23112.467110
    
    

      
std
      
0.835158
      
3.703649
      
2202.493057
      
77859.160373
    
    

      
min
      
3.000000
      
5.000000
      
2.000000
      
0.695000
    
    

      
25%
      
3.000000
      
5.000000
      
6.000000
      
2.810000
    
    

      
50%
      
4.000000
      
10.000000
      
17.000000
      
13.605000
    
    

      
75%
      
5.000000
      
10.000000
      
921.000000
      
4119.600000
    
    

      
max
      
5.000000
      
15.000000
      
9737.000000
      
990358.715000
    
    

      
2
      
count
      
787.000000
      
787.000000
      
787.000000
      
787.000000
    
    

      
mean
      
3.946633
      
9.282084
      
783.714104
      
20018.015305
    
    

      
std
      
0.827146
      
3.869595
      
1878.861366
      
84576.032327
    
    

      
min
      
3.000000
      
5.000000
      
2.000000
      
0.655000
    
    

      
25%
      
3.000000
      
5.000000
      
5.000000
      
1.917500
    
    

      
50%
      
4.000000
      
10.000000
      
10.000000
      
4.460000
    
    

      
75%
      
5.000000
      
10.000000
      
329.000000
      
542.470000
    
    

      
max
      
5.000000
      
15.000000
      
9995.000000
      
875336.620000
    
  

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    

sns.factorplot(x='dimension', y='steps', data=df, hue='rand_degree', col='heuristic', size=7)


    

In [ ]:

    

sns.factorplot(x='dimension', y='time', data=df, hue='rand_degree', col='heuristic', size=7)


    

In [ ]:

    

g = sns.FacetGrid(data=df, row='dimension', col='heuristic', hue='rand_degree', sharex=True, sharey=True, aspect=2, size=7)
g =(g.map(sns.pointplot, "time", "steps").add_legend())


    

In [ ]:

    

sns.stripplot( x="dimension", y="steps", data=df, jitter=True,hue="rand_degree", size=4);


    

In [ ]:

    

sns.stripplot(x="rand_degree", y="time", data=df, jitter=True, hue="dimension", size=4);


    

In [ ]:

    

sns.stripplot(x="dimension", y="steps", data=df, jitter=True, hue="rand_degree", size=4);


    

In [ ]:

    

sns.stripplot(x="dimension", y="time", data=df, jitter=True, hue="rand_degree", size=4);


    

In [ ]:

    

sns.swarmplot(x="dimension", y="steps", data=df, hue="rand_degree");


    

In [ ]:

    

sns.swarmplot(x="dimension", y="time", data=df, hue="rand_degree");


    

In [ ]:

    

sns.swarmplot(x="rand_degree", y="steps", data=df, hue="dimension");


    

In [ ]:

    

sns.swarmplot(x="rand_degree", y="time", data=df, hue="dimension");


    

In [ ]:

    

sns.boxplot(x="rand_degree", y="steps", data=df, hue="dimension");


    

In [ ]:

    

sns.boxplot(x="rand_degree", y="time", data=df, hue="dimension");


    

In [ ]:

    

sns.boxplot(x="dimension", y="steps", data=df, hue="rand_degree");


    

In [ ]:

    

sns.boxplot(x="dimension", y="time", data=df, hue="rand_degree");


    

In [ ]:

    

sns.lmplot(x="steps", y="time", hue="dimension", col="heuristic", row='rand_degree', data=df)


    

In [ ]:

    

sns.lmplot(x="steps", y="time", hue="rand_degree", col="heuristic", row='dimension', data=df)


    

In [ ]:

    

sns.lmplot(x="steps", y="time", row="dimension", col="heuristic", hue="rand_degree",data=df)


    

In [ ]:

    

sns.lmplot(x="steps", y="time", hue="dimension", col="heuristic", row="rand_degree", data=df)


    

In [ ]:

    

sns.distplot(df['time'], bins=3)


    

In [ ]:

    

sns.lmplot(y="time", x="steps", col="dimension", row="rand_degree", data=df)


    

In [ ]:

    

sns.lmplot(y="steps", x="dimension", row="dimension", col="heuristic",hue="rand_degree", data=df)


    

In [ ]:

    

sns.lmplot(y="time", x="dimension", row="dimension", col="heuristic",hue="rand_degree", data=df)


    

In [ ]:

    

sns.lmplot(y="steps", x="dimension", hue="dimension", row="rand_degree", col="heuristic", data=df,  x_jitter=.1)


    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [ ]:

    




    

In [55]:

    

g = sns.lmplot(x="Steps", y="Time in minutes", hue="Rand_degree", col="Heuristic", row='Dimension', data=df)


    

In [44]:

    

sns.factorplot(x='Dimension', y='Steps', data=df, hue='Rand_degree', col='Heuristic', size=7)


    

    
Out[44]:





<seaborn.axisgrid.FacetGrid at 0x11e61c208>






    

In [45]:

    

sns.factorplot(x='Dimension', y='Time in minutes', data=df, hue='Rand_degree', col='Heuristic', size=7)


    

    
Out[45]:





<seaborn.axisgrid.FacetGrid at 0x11fd33be0>






    

In [12]:

    

g = sns.lmplot(y="Steps", x="Dimension", row="Dimension", col="Heuristic",hue="Rand_degree", data=df)
g.set(xticks=[3,4,5])


    

    
Out[12]:





<seaborn.axisgrid.FacetGrid at 0x119f3d240>






    

In [15]:

    

g= sns.lmplot(y="Time in minutes", x="Dimension", row="Dimension", col="Heuristic",hue="Rand_degree", data=df)
g.set(xticks=[3,4,5])


    

    
Out[15]:





<seaborn.axisgrid.FacetGrid at 0x115a9d160>






    

In [48]:

    

sns.boxplot(x="Rand_degree", y="Steps", data=df, hue="Dimension");


    

In [ ]:

    




    

In [ ]: