In [40]:

    

import numpy as np
import scipy as sp
import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
import pymongo
import os

# import branin
import sys,os
sys.path.append(os.path.join(os.getcwd() ,"test_py2"))
from branin import branin
# awesome plot options
plt.style.use('fivethirtyeight')
sns.set(style="white",font_scale=1.5,)
plt.rcParams['figure.figsize'] = (12.0, 6.0)
plt.rcParams['savefig.dpi'] = 100
from matplotlib.colors import LogNorm
%config InlineBackend.figure_format='retina'
%matplotlib inline


    

In [2]:

    

!head -n 50 test_py2/branin.py


    

    




from __future__ import print_function
import numpy as np


def branin(x, y):
    result = np.square(y - (5.1 / (4 * np.square(np.pi))) * np.square(x) +
                       (5 / np.pi) * x - 6) + 10 * (1 - (1. / (8 * np.pi))) * np.cos(x) + 10
    return float(result)


def main(job_id, params):
    print('Job #%d' % job_id)
    print(params)
    result = branin(params['x'], params['y'])
    print('Result = %f' % result)
    return result

In [57]:

    

locmin_z = [3.14159213 , 2.27500065]
min_z = branin(locmin_z[0],locmin_z[1])
print("minima %f at %s "%(min_z,str(locmin_z)))


    

    




minima 0.397887 at [3.14159213, 2.27500065] 

In [58]:

    

x = np.linspace(-5, 10, 500 )
y = np.linspace(0, 15, 500 )
Z = np.zeros((x.shape[0],y.shape[0]))
X, Y = np.meshgrid(x, y)
Z = np.array([[ branin(xx,yy) for yy in y] for xx in x])
        
plt.figure(figsize=(8,6))
plt.pcolor(X,Y,Z.T,
           norm=LogNorm(vmin=np.min(Z)*0.9,vmax=np.max(Z)),cmap='bone')
plt.colorbar(label='$f(x,y)$')
levels = np.logspace(np.log10(np.min(Z)*0.9), np.log10(np.max(Z)), 20)
plt.contour(Z.T, levels,linewidths=2,alpha=0.5,extent=(-5,10,0,15))
plt.scatter(min_x,min_y,s=100,c='g',label='global_minima')
# style
plt.legend(loc='best')
plt.xlabel('$x$')
plt.ylabel('$y$')
sns.despine(trim=True)
plt.show()


    

In [75]:

    

from scipy.optimize import minimize
guess = (np.mean(x),np.mean(y))
methods = {'BFGS':[],'CG':[],'Nelder-Mead':[]}

def info(x):
    pos.append(x)

branin_vector = lambda x: branin(x[0],x[1])
for key,val in methods.items():
    pos =[]
    print('== %s ==='%key)
    res = minimize( branin_vector, guess, jac=False,method=key,callback=info,options={'disp': True})
    methods[key]=np.array(pos)
    print("success? %s "%(res['success']))
    if res['success']:
        print("Global minima %f at %s "%(min_z,str(locmin_z)))
        print("Found minima %f at %s "%(branin(res['x'][0],res['x'][1]),str(res['x'])))


    

    




== Nelder-Mead ===
Optimization terminated successfully.
         Current function value: 0.397887
         Iterations: 41
         Function evaluations: 80
success? True 
Global minima 0.397887 at [3.14159213, 2.27500065] 
Found minima 0.397887 at [ 3.14157157  2.27501556] 
== BFGS ===
Optimization terminated successfully.
         Current function value: 0.397887
         Iterations: 7
         Function evaluations: 40
         Gradient evaluations: 10
success? True 
Global minima 0.397887 at [3.14159213, 2.27500065] 
Found minima 0.397887 at [ 3.14159213  2.27500063] 
== CG ===
Warning: Desired error not necessarily achieved due to precision loss.
         Current function value: 0.397891
         Iterations: 6
         Function evaluations: 56
         Gradient evaluations: 11
success? False 

In [76]:

    

plt.figure(figsize=(8,6))
plt.pcolor(X,Y,Z.T,
           norm=LogNorm(vmin=np.min(Z)*0.9,vmax=np.max(Z)),cmap='bone')
plt.colorbar(label='$f(x,y)$')
levels = np.logspace(np.log10(np.min(Z)*0.9), np.log10(np.max(Z)), 20)
plt.contour(Z.T, levels,linewidths=2,alpha=0.5,extent=(-5,10,0,15))
plt.scatter(min_x,min_y,s=100,c='g',label='global_minima')
for key,val in methods.items():
    plt.plot(val[:,0],val[:,1],'-o',label=key)
    #plt.scatter(val['x'][0],val['x'][1],s=500,c='r',label='Grid')
# style
plt.legend(loc='best')
plt.xlabel('$x$')
plt.ylabel('$y$')
sns.despine(trim=True)
plt.show()


    

In [77]:

    

!mongod --fork --logpath test_py2/db/mongodb.log --dbpath test_py2/db


    

    




about to fork child process, waiting until server is ready for connections.
forked process: 148
ERROR: child process failed, exited with error number 48

In [78]:

    

from pymongo import MongoClient
client = MongoClient()
print("Avail DB's: {}".format(client.database_names()))
db = client['spearmint']
print('DB fields: {}'.format(db.collection_names()))
coll= db['simple-braninhoo-example.jobs']


    

    




Avail DB's: [u'local', u'simple-braninhoo-example', u'spearmint']
DB fields: [u'simple-braninhoo-example.jobs', u'simple-braninhoo-example.recommendations', u'simple-braninhoo-example.hypers']

In [83]:

    

import zlib
def decompress(values):
    val = np.fromstring(zlib.decompress(values['value'].decode('base64'))).reshape(values['shape'])
    if values['shape'] == [1]:
        val = val[0]
    return val

rows = []
for rec in coll.find():
    delta_t = rec['end time']-rec['start time']
    a_dict={'id':rec['id'],'objective':rec['values']['Objective'],'time':delta_t}
    for key,val in rec['params'].items():
        a_dict.update({key:decompress(val['values'])})
    rows.append(a_dict)
df = pd.DataFrame(rows)    
print(df.shape)
df.head()


    

    




(66, 5)






    
Out[83]:






  

    

      

      
id
      
objective
      
time
      
x
      
y
    
  
  

    

      
0
      
1
      
14.711279
      
0.001352
      
9.597626
      
6.389008
    
    

      
1
      
2
      
26.680522
      
0.001381
      
-4.990035
      
13.410486
    
    

      
2
      
3
      
16.901297
      
0.001250
      
4.793785
      
3.813145
    
    

      
3
      
4
      
141.133644
      
0.001311
      
9.952873
      
14.764435
    
    

      
4
      
5
      
252.608179
      
0.001255
      
-4.616085
      
0.493928
    
  

In [ ]: