def get_mf_pca_df(rows=5):
"""
function get_mf_pca_df(rows=5) shows some results
that were predicted for matrix+pca features by all methods
"""
mf_pca_df = load_dataframe('matrix_pca_df')
return mf_pca_df[:rows]
def get_haars_df(rows=5):
"""
function get_haars_df(rows=5) shows some results
that were predicted for haars features by all methods
"""
haars_df = load_dataframe('haars_df')
return haars_df[:rows]
def get_haars_net_df(rows=5, exp_num=1):
"""
function get_haars_net_df(rows=5, exp_num=1) shows some results
that were predicted for haars features by neural networks at
first or second experiment
"""
haars_net_df = load_dataframe('haars_net_df' + '_exp_num_' + str(exp_num))
return haars_net_df[:rows]
def calc_accuracy():
"""
function calc_accuracy() calculate accuracy
and returns dataframe with results
"""
col_list = COL_LIST
col_list.append('voting (cols [1-5) )')
haars_df = load_dataframe('haars_df')
matrix_df = load_dataframe('matrix_df')
matrix_pca_df = load_dataframe('matrix_pca_df')
accuracy_df = pd.DataFrame(FEATURES_LIST, index=['1', '2', '3'], columns=['features'])
for i, cols in enumerate(col_list):
accuracy_df[cols] = [accuracy_score(haars_df['y_test'], haars_df[cols]),
accuracy_score(matrix_df['y_test'], matrix_df[cols]),
accuracy_score(matrix_pca_df['y_test'], matrix_pca_df[cols])]
return accuracy_df[:]
def show_wrong_classified_obj():
"""
function show_wrong_classified_obj() shows some examples
ml-models gave false positive or false negative detections
"""
indexes = np.arange(1, 96, 32)
col_list = COL_LIST
col_list.append('voting (cols [1-5) )')
x_test = load_sample('x_test')
fig = plt.figure(figsize=(8, 60))
ax = []
plt.rcParams.update({'font.size': 6})
np.random.seed(RANDOM_SEED)
plotting(load_dataframe('haars_df'), 'hf', indexes[0],
col_list, fig, ax, x_test)
plotting(load_dataframe('matrix_df'), 'mf', indexes[1],
col_list, fig, ax, x_test)
plotting(load_dataframe('matrix_pca_df'), 'mf_pca', indexes[2],
col_list, fig, ax, x_test)
plt.show()
def get_time_df():
"""
function get_time_df() returns dataframe with time
"""
time_part1_df = load_dataframe('time_df_part1')
return time_part1_df[:]
def get_net_time_df(exp_num=1):
"""
function get_net_time_df(exp_num=1) returns dataframe with time
"""
net_time_df = load_dataframe('net_time_df' + '_exp_num_' + str(exp_num))
return net_time_df[:]