def helper_extract_leading_digits_payments(num_payments, col, num_digits):
prefix = TEST_DATA_DIR + "/payments_" + str(num_payments) + "_"
input_filename = prefix+"input.csv"
amounts = read_column_from_csv(input_filename, col, True)
amounts_copy = amounts[:]
expected_filename = prefix + str(num_digits) + "_leading_digits.txt"
expected = []
for x in read_column_from_csv(expected_filename, 0, True):
expected.append(int(x))
leading_digits = extract_leading_digits_from_list(amounts, num_digits)
assert leading_digits == expected
assert amounts == amounts_copy, "Do not change the list that is passed to your function!"
def go():
usage = "usage: python benford.py <input filename> <column number> <num digits>"
if len(sys.argv) != 4:
print(usage)
else:
input_filename = sys.argv[1]
if not os.path.isfile(input_filename):
print(usage)
print("error: file not found: {}".format(input_filename))
return
# convert column number argument to an integer
try:
col_num = int(sys.argv[2])
except ValueError:
s = "error: column number must be an integer: {}"
print(usage)
print(s.format(sys.argv[2]))
return
data = util.read_column_from_csv(input_filename, col_num, True)
# convert number of digits argument to an integer
try:
num_digits = int(sys.argv[3])
except ValueError:
s = "error: number of digits must be an integer: {}".format(sys.argv[3])
print(usage)
print(s.format(sys.argv[3]))
return
plot_benford_dist(data, num_digits)
# print only four digits after the decimal point
print("MAD: {:.4}".format(compute_benford_MAD(data, num_digits)))
def compare_actual_expected_from_file(actual, expected_filename):
# get expected list of values from the file
expected = []
for x in read_column_from_csv(expected_filename, 0, True):
expected.append(float(x))
if not isinstance(actual, list):
s = "Actual value returned from the function must be a list of floats."
pytest.fail(s)
if len(actual) > 0 and isinstance(actual[0], list):
s = (
"Actual value returned from the function must be a list of floats,"
" not a list containing a list.")
pytest.fail(s)
if len(actual) != len(expected):
s = "Length of expected ({0}) and actual results ({1}) do not match"
pytest.fail(s.format(len(expected), len(actual)))
for i in range(len(actual)):
# stored and computed representations may not be identical
if abs(expected[i] - actual[i]) > EPS:
s = "actual and expected values do not match at element {0}"
pytest.fail(s.format(i))
def helper_test_compute_benford_MAD(prefix, col, num_digits):
input_filename = prefix + "input.csv"
amounts = read_column_from_csv(input_filename, col, True)
actual = compute_benford_MAD(amounts, num_digits)
expected_filename = prefix + "computed_benford_mad_{0}_output.txt".format(num_digits)
compare_actual_expected_from_file([actual], expected_filename)
def compare_actual_expected_from_file(actual, expected_filename):
# get expected list of values from the file
expected = []
for x in read_column_from_csv(expected_filename, 0, True):
expected.append(float(x))
compare_actual_expected(actual, expected)
def helper_test_compute_benford_dist(prefix, col, num_digits):
input_filename = prefix + "input.csv"
amounts = read_column_from_csv(input_filename, col, True)
amounts_copy = amounts[:]
actual = compute_benford_dist(amounts, num_digits)
expected_filename = prefix + "computed_benford_dist_{0}_output.txt".format(num_digits)
compare_actual_expected_from_file(actual, expected_filename)
assert amounts == amounts_copy, "Do not change the list that is passed to your function!"
def helper_test_compute_benford_dist(prefix, col, num_digits):
input_filename = prefix + "input.csv"
amounts = read_column_from_csv(input_filename, col, True)
expected_filename = prefix + "computed_benford_dist_{0}_output.txt".format(
num_digits)
amounts_copy = amounts[:]
actual = compute_benford_dist(amounts, num_digits)
compare_actual_expected_from_file(actual, expected_filename)
if amounts != amounts_copy:
pytest.fail("Do not change the list that is passed to your function!")
def helper_test_compute_benford_dist(prefix, col, currency_symbol, num_digits):
input_filename = os.path.join(TEST_DATA_DIR, prefix + "input.csv")
amounts = read_column_from_csv(input_filename, col, True)
expected_filename = os.path.join(
TEST_DATA_DIR,
prefix + "computed_benford_dist_{0}_output.txt".format(num_digits))
amounts_copy = amounts[:]
actual = benford.compute_benford_dist(currency_symbol, amounts, num_digits)
compare_actual_expected_from_file(actual, expected_filename)
if amounts != amounts_copy:
pytest.fail("Do not change the list that is passed to your function!")
def helper_test_compute_benford_MAD(prefix, col, currency_symbol, num_digits):
input_filename = prefix + "input.csv"
amounts = read_column_from_csv(input_filename, col, True)
amounts_copy = amounts[:]
actual = benford.compute_benford_MAD(currency_symbol, amounts, num_digits)
if amounts != amounts_copy:
pytest.fail("Do not change the list that is passed to your function!")
# get expected list of values from the file
expected_filename = prefix + "computed_benford_mad_{0}_output.txt".format(
num_digits)
expected = float(read_column_from_csv(expected_filename, 0, True)[0])
if not isinstance(actual, float):
s = "Actual value returned from the function must be a float."
pytest.fail(s)
if abs(expected - actual) > EPS:
s = "actual ({:f}) and expected ({:f}) values do not match"
pytest.fail(s.format(actual, expected))
compare_actual_expected_from_file([actual], expected_filename)
def go():
'''
Process the arguments and do the work.
'''
usage = ("usage: python benford.py <input filename> <column number>"
"<currency symbol> <num digits>")
if len(sys.argv) < 5 or len(sys.argv) > 6:
print(usage)
return
input_filename = sys.argv[1]
if not os.path.isfile(input_filename):
print(usage)
print("error: file not found: {}".format(input_filename))
return
# convert column number argument to an integer
try:
col_num = int(sys.argv[2])
except ValueError:
s = "error: column number must be an integer: {}"
print(usage)
print(s.format(sys.argv[2]))
return
data = util.read_column_from_csv(input_filename, col_num, True)
currency_symbol = sys.argv[3]
# convert number of digits argument to an integer
try:
num_digits = int(sys.argv[4])
except ValueError:
s = "error: number of digits must be an integer: {}".format(sys.argv[4])
print(usage)
print(s.format(sys.argv[4]))
return
# grab the name for the PNG file, if exists.
if len(sys.argv) == 5:
output_filename = None
else:
output_filename = sys.argv[5]
plot_benford_dist(currency_symbol, data, num_digits, output_filename)
# print only four digits after the decimal point
print("MAD: {:.4}".format(compute_benford_MAD(currency_symbol, data, num_digits)))
plt.xticks(range(lb, ub, 10**(num_digits-1)))
# compute limits for the y axis
max_val = max(max(expected), max(actual))
y_ub = max_val + max_val * .1
plt.ylim(0,y_ub)
# add labels
plt.title("Actual (blue) and expected (red) Benford distributions")
if num_digits ==1:
plt.xlabel("Leading digit")
else:
plt.xlabel("Leading digits")
plt.ylabel("Proportion")
plt.savefig(output_filename)
if __name__=="__main__":
if len(sys.argv) != 5:
print("usage: python benford.py <input filename> <column number> <num digits> <output filename>")
else:
input_filename = sys.argv[1]
data = util.read_column_from_csv(input_filename, int(sys.argv[2]), True)
num_digits = int(sys.argv[3])
output_filename = sys.argv[4]
plot_benford_dist(data, num_digits, output_filename)
# print only four digits after the decimal point
print("MAD: {0:.4}".format(benford.compute_benford_MAD(data, num_digits)))