def evaluate(model_path=os.path.join('/home/ubuntu/checkpoints',
'xception05-0.236.mod')):
# load test data
test_X = np.load(os.path.join('/home/ubuntu/training_data', 'test_X.npy'))
test_Y = np.load(os.path.join('/home/ubuntu/training_data', 'test_Y.npy'))
# load model
model = model = load_model(model_path)
# get predictions
test_predictions = model.predict(test_X, batch_size=16)
# get measures
test_measures = measures.get_measures(test_predictions, test_Y, .5)
# read genres
genre_file_path = os.path.join('/home/ubuntu/training_data', 'genres.txt')
with open(genre_file_path, 'r') as handler:
genres = handler.readlines()
genres = [genre[:-1] for genre in genres]
# print measures
print("Statistics on test data:")
measures.print_measures(test_measures, genres)
def evaluate(
model_path=os.path.join('model', 'mod'), cutoff_file='cutoffs.npy'):
# load test data
test_X = np.load(os.path.join('training_data', 'test_X.npy'))
test_Y = np.load(os.path.join('training_data', 'test_Y.npy'))
# load model
model = model = load_model(model_path)
# load cutoffs
cutoffs = np.load(os.path.join('cutoffs', cutoff_file))
# get predictions
test_predictions = model.predict(test_X, batch_size=16)
# get measures
test_measures = measures.get_measures(test_predictions, test_Y, cutoffs)
# read genres
genre_file_path = os.path.join('training_data', 'genres.txt')
with open(genre_file_path, 'r') as handler:
genres = handler.readlines()
genres = [genre[:-1] for genre in genres]
# print measures
print("Statistics on test data:")
measures.print_measures(test_measures, genres)
def fit_predict(model, train, test):
"""
fit model on train, predict on test and return accuracy
"""
model.fit(train)
predict, true = model.predict(test)
TP, TN, FP, FN, acc, recall, precision, f1 = get_measures(predict, true)
return acc
def check_knn():
ds_ = DataSet("dataset.txt")
knn_ = Knn(5)
knn_.fit(ds_)
predict, true = knn_.predict(ds_)
TP, TN, FP, FN, acc, recall, precision, f1 = get_measures(predict, true)
print("accuracy:", acc, "\n"
"recall:", recall, "\n"
"precision:", precision, "\n"
"f1:", f1, "\n")
def check_naive_bayes():
ds_ = DataSet("dataset.txt")
naive_bayes_ = NaiveBayes(ds_.header)
naive_bayes_.fit(ds_)
predict, true = naive_bayes_.predict(ds_)
TP, TN, FP, FN, acc, recall, precision, f1 = get_measures(predict, true)
print("accuracy:", acc, "\n"
"recall:", recall, "\n"
"precision:", precision, "\n"
"f1:", f1, "\n")
def test_scalability(timeout=10):
results = defaultdict(list)
nodes = [100, 1000, 10000, 100000, 1000000]
graphs = [graph_loader(graph_type='CSF', n=n, seed=1) for n in nodes]
measures = get_measures()
measures = measures + graph_approx + spectral_approx
for measure in tqdm(measures):
run_times = parallelize_evaluation(graphs, measure, timeout)
results[measure] = run_times
color = plt.cm.gist_ncar(np.linspace(0, 0.9, len(measures)))
mpl.rcParams['axes.prop_cycle'] = cycler.cycler('color', color)
mpl.rcParams['legend.fontsize'] = 5
fig, axes = plt.subplots(ncols=5, figsize=(5 * 6 - 1, 5))
for (measure, run_time) in results.items():
axes[0].plot(nodes,
run_time,
label=measure,
linewidth=1,
linestyle=measure_style[measure])
axes[0].set_xlabel('Number of nodes')
axes[0].set_ylabel('Time in seconds')
axes[1].plot(nodes,
run_time,
label=measure,
linewidth=1,
linestyle=measure_style[measure])
axes[1].set_xscale('log')
axes[1].set_xlabel('Number of nodes (log-scale)')
axes[1].set_ylabel('Time in seconds')
axes[2].plot(nodes,
run_time,
label=measure,
linewidth=1,
linestyle=measure_style[measure])
axes[2].set_xscale('log')
axes[2].set_yscale('log')
axes[2].set_xlabel('Number of nodes (log-scale)')
axes[2].set_ylabel('Time in seconds (log-scale)')
axes[0].legend()
plt.title('Clustered Scale Free Graph')
save_dir = os.getcwd() + '/plots/'
os.makedirs(save_dir, exist_ok=True)
plt.savefig(save_dir + 'scalability.pdf')
def check_dtl():
ds_ = DataSet("dataset.txt")
dtl_ = DecisionTree(ds_.header)
tree_ = dtl_.fit(ds_)
predict, true = dtl_.predict(ds_)
TP, TN, FP, FN, acc, recall, precision, f1 = get_measures(predict, true)
f = open("tree.txt", "wt")
f.write(str(dtl_))
print("accuracy:", acc, "\n"
"recall:", recall, "\n"
"precision:", precision, "\n"
"f1:", f1, "\n")
e = 0