Support vector machine (SVM) calibration experiment

In [6]:

    

from sklearn import grid_search
from sklearn.cross_validation import StratifiedShuffleSplit
from sklearn.cross_validation import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.svm import SVC
from time import time

import numpy as np
import pandas as pd
import sys


    

In [10]:

    

#Test each array of features in the training dataset
if __name__ == '__main__':
    
    csv_file_features = [[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13], 
                         [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 13]]
    
    estimators_array = []
    
    print('Order of results')
    print('Training score, Test score, Time, Features, SVM parameters\n')
    
    C_range = [0.01, 0.1, 1, 10, 100]
    gamma_range = [0.01, 0.1, 1, 10, 100]

    #Test for all the features sets
    for i in csv_file_features:
        stdsc = StandardScaler()
        df = pd.read_csv("../../Dataset/Train/EEG_Train_Sorted.csv")

        #Separate Class labels anda data
        X = df.ix[:, i]
        y = df['Class']
        
        X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=0)
        
        #Standarize data
        X_train_std = stdsc.fit_transform(X_train)
        X_test_std = stdsc.transform(X_test)
        
        #Create SVM and define parameters
        svm = SVC()
        param_grid = [{'C': C_range, 
                        'gamma': gamma_range, 
                        'kernel': ['rbf'], 
                        'decision_function_shape': ['ovr'], 
                        'random_state': [0]}]
            
        #Set Grid Params
        init = time()
        gscv = grid_search.GridSearchCV(svm, 
                                       param_grid, 
                                       n_jobs=6, 
                                       cv=StratifiedShuffleSplit(y=y_train,  
                                                                 n_iter=5, 
                                                                 test_size=0.4, 
                                                                 random_state=0))
        #Test classifier
        gscv.fit(X_train_std, y_train)
        sys.exit("Error message")
        #Save our estimator
        estimators_array.append(gscv)
        
        #Report testing results
        print('{:.5f}, {:.5f}, {:.5f}s, {}, {}'.format(gscv.best_score_, gscv.score(X_test_std, y_test), time() - init, i, gscv.best_params_))


    

    




Order of results
Training score, Test score, Time, Features, SVM parameters







    




C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\hashing.py:197: DeprecationWarning: Changing the shape of non-C contiguous array by
descriptor assignment is deprecated. To maintain
the Fortran contiguity of a multidimensional Fortran
array, use 'a.T.view(...).T' instead
  obj_bytes_view = obj.view(self.np.uint8)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\pool.py:436: UserWarning: Failed to clean temporary folder: C:\Users\c_am_\AppData\Local\Temp\joblib_memmaping_pool_7456_1997645054584
  warnings.warn("Failed to clean temporary folder: %s" % folder_path)






    




---------------------------------------------------------------------------
KeyboardInterrupt                         Traceback (most recent call last)
C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\parallel.py in retrieve(self)
    726             try:
--> 727                 self._output.extend(job.get())
    728             except tuple(self.exceptions) as exception:

C:\Anaconda3\lib\multiprocessing\pool.py in get(self, timeout)
    601     def get(self, timeout=None):
--> 602         self.wait(timeout)
    603         if not self.ready():

C:\Anaconda3\lib\multiprocessing\pool.py in wait(self, timeout)
    598     def wait(self, timeout=None):
--> 599         self._event.wait(timeout)
    600 

C:\Anaconda3\lib\threading.py in wait(self, timeout)
    548             if not signaled:
--> 549                 signaled = self._cond.wait(timeout)
    550             return signaled

C:\Anaconda3\lib\threading.py in wait(self, timeout)
    292             if timeout is None:
--> 293                 waiter.acquire()
    294                 gotit = True

KeyboardInterrupt: 

During handling of the above exception, another exception occurred:

KeyboardInterrupt                         Traceback (most recent call last)
<ipython-input-10-65307e56070b> in <module>()
     46                                                                  random_state=0))
     47         #Test classifier
---> 48         gscv.fit(X_train_std, y_train)
     49         sys.exit("Error message")
     50         #Save our estimator

C:\Anaconda3\lib\site-packages\sklearn\grid_search.py in fit(self, X, y)
    802 
    803         """
--> 804         return self._fit(X, y, ParameterGrid(self.param_grid))
    805 
    806 

C:\Anaconda3\lib\site-packages\sklearn\grid_search.py in _fit(self, X, y, parameter_iterable)
    551                                     self.fit_params, return_parameters=True,
    552                                     error_score=self.error_score)
--> 553                 for parameters in parameter_iterable
    554                 for train, test in cv)
    555 

C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\parallel.py in __call__(self, iterable)
    808                 # consumption.
    809                 self._iterating = False
--> 810             self.retrieve()
    811             # Make sure that we get a last message telling us we are done
    812             elapsed_time = time.time() - self._start_time

C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\parallel.py in retrieve(self)
    748                 # the results as we will raise the exception we got back
    749                 # to the caller instead of returning any result.
--> 750                 self._terminate_pool()
    751                 if self._managed_pool:
    752                     # In case we had to terminate a managed pool, let

C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\parallel.py in _terminate_pool(self)
    547         if self._pool is not None:
    548             self._pool.close()
--> 549             self._pool.terminate()  # terminate does a join()
    550             self._pool = None
    551             if self.backend == 'multiprocessing':

C:\Anaconda3\lib\site-packages\sklearn\externals\joblib\pool.py in terminate(self)
    581 
    582     def terminate(self):
--> 583         super(MemmapingPool, self).terminate()
    584         delete_folder(self._temp_folder)

C:\Anaconda3\lib\multiprocessing\pool.py in terminate(self)
    503         self._state = TERMINATE
    504         self._worker_handler._state = TERMINATE
--> 505         self._terminate()
    506 
    507     def join(self):

C:\Anaconda3\lib\multiprocessing\util.py in __call__(self, wr, _finalizer_registry, sub_debug, getpid)
    183                 sub_debug('finalizer calling %s with args %s and kwargs %s',
    184                           self._callback, self._args, self._kwargs)
--> 185                 res = self._callback(*self._args, **self._kwargs)
    186             self._weakref = self._callback = self._args = \
    187                             self._kwargs = self._key = None

C:\Anaconda3\lib\multiprocessing\pool.py in _terminate_pool(cls, taskqueue, inqueue, outqueue, pool, worker_handler, task_handler, result_handler, cache)
    544         util.debug('joining worker handler')
    545         if threading.current_thread() is not worker_handler:
--> 546             worker_handler.join()
    547 
    548         # Terminate workers which haven't already finished.

C:\Anaconda3\lib\threading.py in join(self, timeout)
   1052 
   1053         if timeout is None:
-> 1054             self._wait_for_tstate_lock()
   1055         else:
   1056             # the behavior of a negative timeout isn't documented, but

C:\Anaconda3\lib\threading.py in _wait_for_tstate_lock(self, block, timeout)
   1068         if lock is None:  # already determined that the C code is done
   1069             assert self._is_stopped
-> 1070         elif lock.acquire(block, timeout):
   1071             lock.release()
   1072             self._stop()

KeyboardInterrupt: 

In [ ]:

    

#Get best estimator parameters
print (estimators_array[0])