def create_bining_strategy_test(function_name):
arg_dict = {
'test': DEFAULT_TEST,
'train': DEFAULT_TRAIN,
'structure': DEFAULT_STRUCTURE,
'tolorance': DEFAULT_TOLORANCE,
'k': DEFAULT_K,
'number_of_bins': DEFAULT_NUMBER_OF_BINS,
'bin_type': DEFAULT_BIN_TYPE,
'missing_values': DEFAULT_MISSING_VALUES,
'8020': DEFAULT_8020
}
resuluts = []
for strategy in ['equal_width', 'equal_frequency', 'entropy']:
result = get_result(function_name, **{
**arg_dict,
**{
'bin_type': strategy
}
})
result_tupple = (function_name, result['score'], result['TP'],
result['TN'], result['FP'], result['FN'], strategy)
resuluts.append(result_tupple)
columns = ('function', 'score', 'TP', 'TN', 'FP', 'FN', 'strategy')
return matrix_to_df(columns, resuluts)
def create_nans_test(function_name):
arg_dict = {
'test': DEFAULT_TEST,
'train': DEFAULT_TRAIN,
'structure': DEFAULT_STRUCTURE,
'tolorance': DEFAULT_TOLORANCE,
'k': DEFAULT_K,
'number_of_bins': DEFAULT_NUMBER_OF_BINS,
'bin_type': DEFAULT_BIN_TYPE,
'missing_values': DEFAULT_MISSING_VALUES,
'8020': DEFAULT_8020
}
resuluts = []
for s in ('remove_nans', 'replace_nans'):
result = get_result(function_name, **{
**arg_dict,
**{
'missing_values': s
}
})
result_tupple = (function_name, result['score'], result['TP'],
result['TN'], result['FP'], result['FN'], s)
resuluts.append(result_tupple)
columns = ('function', 'score', 'TP', 'TN', 'FP', 'FN',
'dealing with nans strategy')
return matrix_to_df(columns, resuluts)
def create_bins_number_test(function_name, bins_array, strategy):
arg_dict = {
'test': DEFAULT_TEST,
'train': DEFAULT_TRAIN,
'structure': DEFAULT_STRUCTURE,
'tolorance': DEFAULT_TOLORANCE,
'k': DEFAULT_K,
'number_of_bins': DEFAULT_NUMBER_OF_BINS,
'bin_type': strategy,
'missing_values': DEFAULT_MISSING_VALUES,
'8020': DEFAULT_8020
}
resuluts = []
for b in bins_array:
print(b)
result = get_result(function_name, **{
**arg_dict,
**{
'number_of_bins': b
}
})
result_tupple = (function_name, result['score'], result['TP'],
result['TN'], result['FP'], result['FN'], b, strategy)
resuluts.append(result_tupple)
columns = ('function', 'score', 'TP', 'TN', 'FP', 'FN', 'number of bins',
'strategy')
return matrix_to_df(columns, resuluts)
def update_output(**kwargs):
setStatus('Working...')
function_name = kwargs['choosen_function_name']
### sending kwargs to start module
### vvvvvvvvvvvv
result = get_result(function_name, **kwargs)
### ^^^^^^^^^^^^^
setScore(result['score'])
output_components['TP_box'].config(text=result['TP'])
output_components['TN_box'].config(text=result['TN'])
output_components['FP_box'].config(text=result['FP'])
output_components['FN_box'].config(text=result['FN'])
setStatus('Done')
def create_k_test(function_name, k_array):
arg_dict = {
'test': DEFAULT_TEST,
'train': DEFAULT_TRAIN,
'structure': DEFAULT_STRUCTURE,
'tolorance': DEFAULT_TOLORANCE,
'k': DEFAULT_K,
'number_of_bins': DEFAULT_NUMBER_OF_BINS,
'bin_type': DEFAULT_BIN_TYPE,
'missing_values': DEFAULT_MISSING_VALUES,
'8020': DEFAULT_8020
}
resuluts = []
for k in k_array:
result = get_result(function_name, **{**arg_dict, **{'k': k}})
result_tupple = (result['score'], result['TP'], result['TN'],
result['FP'], result['FN'], k)
resuluts.append(result_tupple)
columns = ('score', 'TP', 'TN', 'FP', 'FN', 'k')
return matrix_to_df(columns, resuluts)
# function to open structure file
def load_structure():
returnList = []
filename = 'Structure.txt'
try:
tmp = open(filename, "r")
for line in tmp:
returnList.append(line)
tmp.close()
except FileNotFoundError as e:
print("Error", "file was not found!!")
return returnList
# the settings for the single test
kwargs = {
'test': pd.read_csv('test.csv'),
'train': pd.read_csv('train.csv'),
'structure': load_structure(),
'number_of_bins': 14,
'k': 5,
'tolorance': 5,
'bin_type': 'equal_frequency',
'missing_values': 'remove_nans',
'8020': 'yes',
}
# test run example:
print(get_result('id3', **kwargs))