#!/usr/bin/env python
# coding: utf-8

# ## Performance of the optimize function from scipy
# 
# This notebook measures the performance of a simple optimize use case.

# In[24]:


import numpy as np
from scipy.optimize import minimize


# In[25]:


def rosen(x):
    """The Rosenbrock function"""
    return 100.0 * (x[1]-x[0]**2.0)**2.0 + (1-x[0])**2.0


# In[26]:


get_ipython().run_cell_magic('timeit', '', 'minimize(rosen, [0.0, 0.0], method="nelder-mead",\n               options={\'xtol\': 1e-8, \'disp\': False})\n')


# In[27]:


import datetime
print(datetime.datetime.now())
for i in range(1000):
    minimize(rosen, [0.0, 0.0], method="nelder-mead",
               options={'xtol': 1e-8, 'disp': False})
print(datetime.datetime.now())


# ## Let's try numba!

# In[28]:


from numba import jit


# In[29]:


@jit
def rosen2(x):
    """The Rosenbrock function"""
    return 100.0 * (x[1]-x[0]**2.0)**2.0 + (1-x[0])**2.0


# In[30]:


get_ipython().run_cell_magic('timeit', '', 'minimize(rosen2, [0.0, 0.0], method="nelder-mead",\n               options={\'xtol\': 1e-8, \'disp\': False})\n')


# In[31]:


## mmm... it didn't seem to help.


# In[32]:


import numba as nb


# In[33]:


nb.__version__


# In[34]:


import sys


# In[36]:


sys.version


# In[ ]: