def run(create_figures=False, until=42 * 42):
print("Hello world!")
random.seed(1)
env = simpy.Environment()
# the team
team_skills = {
"specify": simpy.Resource(env, 2),
"build": simpy.Resource(env, 2),
"verify": simpy.Resource(env, 2)
}
env.process(work_generator(env, team_skills))
env.run(until=until)
print("Simulation done.")
df = metrics.cycle_times(activities, time_stamps)
# save cycle times to csv file:
df.to_csv(output_cycle_time_file_name, index=False)
if create_figures:
# create cycle time figures
cycle_time_figures, (ax1, ax2) = plt.subplots(nrows=2)
cycle_time_figures.set_size_inches(8.27, 11.69) # A4
ax1 = metrics.cycle_time_scatter_plot(df, ax=ax1)
ax2 = metrics.cycle_time_histogram_plot(df, ax=ax2)
cycle_time_figures.savefig('./output/02_cycle_time_plots.pdf')
def test_retrieve_event_stream_from_cycle_time_dataframe():
timestamps = {'a': [1, 2, 3, 5, 7], 'b': [2.5, 5.7, 6.5, 8.8, 9.1]}
cycle_times_df = metrics.cycle_times(['Build', 'Verify'], timestamps)
stream = metrics.get_event_stream(cycle_times_df)
assert len(stream) == 8
assert stream[0] == 2 # first timestamp for a.build_started
def test_cfd_dataframe_has_zeros_first_row():
timestamps = {'a': [1, 2, 3, 5, 7], 'b': [2.5, 5.7, 6.5, 8.8, 9.1]}
cycle_times_df = metrics.cycle_times(['Build', 'Verify'], timestamps)
df = metrics.get_cfd_dataframe(cycle_times_df)
first_row_with_counts = df.iloc[0, 1:]
assert np.all(first_row_with_counts == 0)
def test_started_and_completed_column_added_and_set_to_first_started_timestamp(
):
timestamps = {'a user story': [1, 2, 3]}
df = metrics.cycle_times(['Build'], timestamps)
assert 'Started' in df.columns
assert 'Completed' in df.columns
assert df['Started']['a user story'] == 2
assert df['Completed']['a user story'] == 3
def test_cycle_time_flow_columns_are_at_known_location():
timestamps = {}
cycle_times_df = metrics.cycle_times(['Build', 'Verify'], timestamps)
cols = metrics.get_flow_columns(cycle_times_df)
assert len(cols) == 4
assert cols[0] == 'Build_started'
assert cols[3] == 'Verify_completed'
def test_cfd_dataframe_everything_completed_at_t_last():
timestamps = {'a': [1, 2, 3, 5, 7], 'b': [2.5, 5.7, 6.5, 8.8, 9.1]}
cycle_times_df = metrics.cycle_times(['Build', 'Verify'], timestamps)
df = metrics.get_cfd_dataframe(cycle_times_df)
# at t_last everything is completed and all counts should be two
tl_expected_counts = np.array([2, 2, 2, 2])
tl_row_with_counts = df.iloc[-1, 1:]
assert (tl_expected_counts == tl_row_with_counts).all()
def test_cfd_dataframe_at_t5():
timestamps = {'a': [1, 2, 3, 5, 7], 'b': [2.5, 5.7, 6.5, 8.8, 9.1]}
cycle_times_df = metrics.cycle_times(['Build', 'Verify'], timestamps)
df = metrics.get_cfd_dataframe(cycle_times_df)
# at t5, task a has started verify stage, but not yet completed
# at t5, task b has not started, so the cunts should be
t5_expected_counts = np.array([1, 1, 1, 0])
t5_row_with_counts = df.iloc[3, 1:]
print(t5_expected_counts)
print(t5_row_with_counts)
assert (t5_expected_counts == t5_row_with_counts).all()
def test_percentiles():
timestamps = {}
for i in range(1, 101):
timestamps[i] = [0, 0, i]
assert len(timestamps) == 100
df = metrics.cycle_times(['Build'], timestamps)
assert df['Cycle_time'][1] == 1
assert df['Cycle_time'][100] == 100
percentile_values = metrics.get_percentile_values(df, (70, 85, 95))
assert abs(percentile_values[0] - 70.3) < 0.1
assert abs(percentile_values[1] - 85.2) < 0.1
assert abs(percentile_values[2] - 95.1) < 0.1
def test_returns_labels_as_column_headers():
df = metrics.cycle_times(['Specify', 'Build', 'Verify'], {})
assert 'Specify_started' in df.columns
assert 'Build_completed' in df.columns
def test_returns_dataframe_from_labels_and_dictionary():
assert isinstance(metrics.cycle_times(['Build'], {}), pd.DataFrame)
def test_cycle_times_column_added():
timestamps = {'a user story': [1, 2, 3], 'a certain bug': [2, 3, 7]}
df = metrics.cycle_times(['Build'], timestamps)
assert 'Cycle_time' in df.columns
assert df['Cycle_time']['a user story'] == 1 # 3-2
assert df['Cycle_time']['a certain bug'] == 4 # 7-3
def test_missing_timestamps_is_OK():
timestamps = {1: [1, 2, 3], 2: [2]}
df = metrics.cycle_times(['Build'], timestamps)
assert len(df) == len(timestamps)
def test_each_value_in_dictionary_results_in_a_row():
timestamps = {1: [1, 2, 3], 2: [2, 3, 4]}
df = metrics.cycle_times(['Build'], timestamps)
assert len(df) == len(timestamps)
def test_starts_with_arrival_timestamp():
df = metrics.cycle_times(['Specify', 'Build', 'Verify'], {})
assert df.columns[0] == 'Arrival'