In [3]:
import pandas as pd
import numpy as np
from sklearn.externals import joblib
from sklearn.pipeline import Pipeline
from sklearn.svm import LinearSVC, SVC
from sklearn.multiclass import OneVsRestClassifier
from sklearn.model_selection import StratifiedShuffleSplit
from sklearn.ensemble import RandomForestClassifier
import matplotlib.pyplot as plt
from scikitplot import classifier_factory
In [4]:
labels = pd.read_csv('./labels.csv', index_col=0)
pts = pd.read_csv('./pts.csv', index_col=0)
In [3]:
pts.info()
In [5]:
all_data = pts.drop([
'matrix_id',
'system_id',
'time',
'matrix_id',
'status_id'], axis=1)
In [ ]:
# plain_data = pts.drop([
# 'matrix_id',
# 'status_id',
# 'time',
# 'HPL_Tflops',
# 'StarDGEMM_Gflops',
# 'SingleDGEMM_Gflops',
# 'PTRANS_GBs',
# 'MPIRandomAccess_LCG_GUPs',
# 'MPIRandomAccess_GUPs',
# 'StarRandomAccess_LCG_GUPs',
# 'SingleRandomAccess_LCG_GUPs',
# 'StarRandomAccess_GUPs',
# 'SingleRandomAccess_GUPs',
# 'StarSTREAM_Copy',
# 'StarSTREAM_Scale',
# 'StarSTREAM_Add',
# 'StarSTREAM_Triad',
# 'SingleSTREAM_Copy',
# 'SingleSTREAM_Scale',
# 'SingleSTREAM_Add',
# 'SingleSTREAM_Triad',
# 'StarFFT_Gflops',
# 'SingleFFT_Gflops',
# 'MPIFFT_Gflops',
# 'MaxPingPongLatency_usec',
# 'RandomlyOrderedRingLatency_usec',
# 'MinPingPongBandwidth_GBytes',
# 'NaturallyOrderedRingBandwidth_GBytes',
# 'RandomlyOrderedRingBandwidth_GBytes',
# 'MinPingPongLatency_usec',
# 'AvgPingPongLatency_usec',
# 'MaxPingPongBandwidth_GBytes',
# 'AvgPingPongBandwidth_GBytes',
# 'NaturallyOrderedRingLatency_usec',
# 'MemProc',
# 'core_count',
# 'cpu_freq',
# 'bogo_mips',
# 'l1_cache',
# 'l2_cache',
# 'l3_cache',
# 'memory_size',
# 'memory_freq',
# 'memory_type'], axis=1)
In [ ]:
all_data.info()
In [ ]:
summit_only_data = plain_data[plain_data['system_id'] == 3]
summit_only_labels = labels[plain_data['system_id'] == 3]
In [ ]:
summit_only_data.info()
summit_only_labels.info()
In [6]:
X = all_data.as_matrix()
y = labels.as_matrix()
In [ ]:
y = y[:,0]
y.transpose()
In [7]:
from sklearn.externals import joblib
classifier = RandomForestClassifier()
classifier = classifier.fit(X,y)
joblib.dump(classifier, 'all_data_classifier.pkl')
Out[7]:
In [10]:
y.shape
Out[10]:
In [15]:
classifiers_list = []
for i in range(0,18):
classifier = RandomForestClassifier()
classifier.fit(X,y[:,i])
#classifiers_list.append(classifier)
classifier_factory(classifier)
classifier.plot_confusion_matrix(X, y[:,i],
cv=10,
title=str(i),
shuffle=True,
figsize=[8,8]
)
plt.show()
#joblib.dump(classifiers_list, 'individual_label_classifiers.pkl')
In [16]:
classifiers_list = []
for i in range(0,18):
classifier = RandomForestClassifier()
classifier.fit(X,y[:,i])
#classifiers_list.append(classifier)
classifier_factory(classifier)
classifier.plot_precision_recall_curve(X, y[:,i],
cv=10,
title=str(i),
shuffle=True,
figsize=[8,8]
)
plt.show()
#joblib.dump(classifiers_list, 'individual_label_classifiers.pkl')
In [17]:
classifiers_list = []
for i in range(0,18):
classifier = RandomForestClassifier()
classifier.fit(X,y[:,i])
#classifiers_list.append(classifier)
classifier_factory(classifier)
classifier.plot_learning_curve(X, y[:,i],
cv=10,
title=str(i),
shuffle=True,
figsize=[8,8]
)
plt.show()
#joblib.dump(classifiers_list, 'individual_label_classifiers.pkl')
In [ ]:
classifiers_list = []
for i in range(0,18):
classifier = RandomForestClassifier()
classifier.fit(X,y[:,i])
#classifiers_list.append(classifier)
classifier_factory(classifier)
classifier.plot_feature_importances(X, y[:,i],
cv=10,
title=str(i),
shuffle=True,
figsize=[8,8]
)
plt.show()
#joblib.dump(classifiers_list, 'individual_label_classifiers.pkl')
In [ ]:
classifiers_list = []
for i in range(0,18):
classifier = RandomForestClassifier()
classifier.fit(X,y[:,i])
#classifiers_list.append(classifier)
classifier_factory(classifier)
classifier.plot_ks_statistic(X, y[:,i],
cv=10,
title=str(i),
shuffle=True,
figsize=[8,8]
)
plt.show()
#joblib.dump(classifiers_list, 'individual_label_classifiers.pkl')
In [7]:
classifiers_list = joblib.load('individual_label_classifiers.pkl')
In [11]:
for i in range (0,18):
cur_y = y[:,i]
pred_y = classifiers_list[i].pred
# classifiers_list[i].plot_confusion_matrix(X, y[:,i],
# cv=10,
# title=str(i),
# shuffle=True)
plt.show()
In [6]:
classifier.plot_precision_recall_curve(X,y[:,0],
shuffle=True,
cv=10,
curves='each_class')
plt.show()
In [ ]:
In [ ]:
classifier = RandomForestClassifier()
classifier_factory(classifier)
sss = StratifiedShuffleSplit()
y_preds = []
y_tests = []
i = 0
for train_index, test_index in sss.split(X, y):
X_train, X_test = X[train_index], X[test_index]
y_train, y_test = y[train_index], y[test_index]
classifier.fit(X_train, y_train)
classifier.plot_roc_curve(X_test,y_test)
plt.show()
y_preds.append(classifier.predict(X_test))
y_tests.append(y_test)
i+=1
In [ ]:
j = i
for i in range(0,j):
y_preds[i] = y_preds[i].astype(int)
y_preds
In [ ]:
from sklearn.metrics import classification_report, confusion_matrix
for i in range(0,10):
print(classification_report(y_tests[i],y_preds[i]))
In [ ]:
from sklearn.metrics import classification_report, confusion_matrix
for i in range(0,10):
print(classification_report(y_tests[i],y_preds[i]))
In [ ]:
from sklearn.model_selection import KFold
classifier = RandomForestClassifier()
kf = KFold(n_splits=10)
y_preds = []
y_tests = []
i = 0
for train_index, test_index in kf.split(X, y):
X_train, X_test = X[train_index], X[test_index]
y_train, y_test = y[train_index], y[test_index]
classifier.fit(X_train, y_train)
y_preds.append(classifier.predict(X_test))
y_tests.append(y_test)
i+=1
In [ ]:
from sklearn.metrics import classification_report, confusion_matrix
for i in range(0,10):
print(classification_report(y_tests[i],y_preds[i]))
In [ ]:
from sklearn.metrics import coverage_error, label_ranking_average_precision_score, label_ranking_loss
print(coverage_error(y_tests[0], y_preds[0]))
print(coverage_error(y_tests[1], y_preds[0]))
print(coverage_error(y_tests[0], y_preds[0]))
In [ ]:
y_pred = classifier.predict(X)
y_pred = y_pred.astype(int)
y_pred
In [ ]:
y_pred_prob = classifier.predict_proba(X)
y_pred_prob
In [ ]:
print("Predicted number of instances:\t",
sum(y_pred[:,0]),
sum(y_pred[:,1]),
sum(y_pred[:,2]),
sum(y_pred[:,3]),
sum(y_pred[:,4]),
sum(y_pred[:,5]))
print("Predicted probabilities of instances:\n",
sum(y_pred_prob[0]), '\n',
sum(y_pred_prob[1]),'\n',
sum(y_pred_prob[2]),'\n',
sum(y_pred_prob[3]),'\n',
sum(y_pred_prob[4]),'\n',
sum(y_pred_prob[5]))
print("Actual number of instances:\t",
sum(y[:,0]),
sum(y[:,1]),
sum(y[:,2]),
sum(y[:,3]),
sum(y[:,4]),
sum(y[:,5]))
In [ ]:
classifier.score(X,y)
In [ ]:
from sklearn.metrics import coverage_error, label_ranking_average_precision_score, label_ranking_loss
print(coverage_error(y, results_y))
print(label_ranking_average_precision_score(y, results_y))
print(label_ranking_loss(y, results_y))
In [ ]:
from sklearn.metrics import confusion_matrix, f1_score, fbeta_score
print(f1_score(y, y_pred, average=None))
print(f1_score(y, y_pred, average='micro'))
print(f1_score(y, y_pred, average='macro'))
print(f1_score(y, y_pred, average='samples'))
print(f1_score(y, y_pred, average='weighted'))
scores = fbeta_score(y, y_pred, beta=0.5, average=None)
In [ ]:
from sklearn.metrics import classification_report, confusion_matrix
print(classification_report(y,y_pred))
print(confusion_matrix(y[:,0],y_pred[:,0]))
In [ ]:
from sklearn.metrics import roc_auc_score, roc_curve
print(roc_auc_score(y, y_pred))
In [ ]:
from sklearn.metrics import zero_one_loss
print(zero_one_loss(y,y_pred)) # Fraction of misclassifications
print(zero_one_loss(y,y_pred,normalize=False)) # num of misclassifications