openai / safety-starter-agents - Conditional Complexity

Intro

Conditional complexity (also called cyclomatic complexity) is a term used to measure the complexity of software. The term refers to the number of possible paths through a program function. A higher value ofter means higher maintenance and testing costs (infosecinstitute.com).
Conditional complexity is calculated by counting all conditions in the program that can affect the execution path (e.g. if statement, loops, switches, and/or operators, try and catch blocks...).
Conditional complexity is measured at the unit level (methods, functions...).
Units are classified in four categories based on the measured McCabe index: 1-5 (simple units), 6-10 (medium complex units), 11-25 (complex units), 26+ (very complex units).

Learn more...

Conditional Complexity Overall

There are 120 units with 1,507 lines of code in units (85.2% of code).

0 very complex units (0 lines of code)
3 complex units (520 lines of code)
4 medium complex units (155 lines of code)
6 simple units (126 lines of code)
107 very simple units (706 lines of code)

Legend:

51+

26-50

11-25

6-10

1-5

Alternative Visuals

3D view of all units

Conditional Complexity per Extension

51+

26-50

11-25

6-10

1-5

Conditional Complexity per Logical Component

primary logical decomposition

51+

26-50

11-25

6-10

1-5

Most Complex Units

Top 50 most complex units

Unit	# lines	McCabe index	# params
def run_polopt_agent() in safe_rl/pg/run_agent.py	241	47	2
def sac() in safe_rl/sac/sac.py	208	27	4
def plot_data() in scripts/plot.py	71	26	11
def get_all_datasets() in scripts/plot.py	29	19	4
def convert_json() in safe_rl/utils/serialization_utils.py	18	13	1
def update_pi() in safe_rl/pg/agents.py	87	12	2
def load_policy() in safe_rl/utils/load_utils.py	21	11	3
def get_datasets() in scripts/plot.py	33	8	2
def compute_gradients() in safe_rl/utils/mpi_tf.py	18	7	4
def dump_tabular() in safe_rl/utils/logx.py	22	7	1
def log_tabular() in safe_rl/utils/logx.py	14	7	5
def update_pi() in safe_rl/pg/agents.py	29	7	2
def _tf_simple_save() in safe_rl/utils/logx.py	10	6	2
def assign_params_from_flat() in safe_rl/utils/mpi_tf.py	5	5	2
def setup_logger_kwargs() in safe_rl/utils/run_utils.py	14	5	4
def mlp_actor_critic() in safe_rl/pg/network.py	14	5	4
def assign_params_from_flat() in safe_rl/pg/trust_region.py	5	5	2
def make_plots() in scripts/plot.py	17	5	16
def main() in scripts/experiment.py	38	5	6
def get_target_update() in safe_rl/sac/sac.py	12	4	3
def mpi_fork() in safe_rl/utils/mpi_tools.py	16	4	2
def mpi_statistics_scalar() in safe_rl/utils/mpi_tools.py	11	4	2
def __init__() in safe_rl/utils/logx.py	17	4	4
def save_state() in safe_rl/utils/logx.py	9	4	3
def ensure_satisfiable_penalty_use() in safe_rl/pg/agents.py	11	4	1
def cg() in safe_rl/pg/trust_region.py	14	4	3
def get_vars() in safe_rl/sac/sac.py	2	3	1
def mpi_op() in safe_rl/utils/mpi_tools.py	6	3	2
def colorize() in safe_rl/utils/logx.py	7	3	4
def restore_tf_graph() in safe_rl/utils/logx.py	12	3	2
def save_config() in safe_rl/utils/logx.py	10	3	2
def setup_tf_saver() in safe_rl/utils/logx.py	4	3	4
def store() in safe_rl/utils/logx.py	5	3	2
def get_stats() in safe_rl/utils/logx.py	4	3	2
def update_pi() in safe_rl/pg/agents.py	11	3	2
def combined_shape() in safe_rl/pg/utils.py	4	3	2
def get() in safe_rl/pg/buffer.py	10	3	1
def placeholder_from_space() in safe_rl/pg/network.py	6	3	1
def get_vars() in safe_rl/pg/network.py	2	3	1
def placeholder() in safe_rl/sac/sac.py	2	2	1
def placeholders() in safe_rl/sac/sac.py	2	2	1
def mlp() in safe_rl/sac/sac.py	4	2	2
def count_vars() in safe_rl/sac/sac.py	3	2	1
def flat_concat() in safe_rl/utils/mpi_tf.py	2	2	1
def apply_gradients() in safe_rl/utils/mpi_tf.py	5	2	4
def log() in safe_rl/utils/logx.py	3	2	3
def log_tabular() in safe_rl/utils/logx.py	7	2	3
def ensure_satisfiable_optimization() in safe_rl/pg/agents.py	6	2	1
def cares_about_cost() in safe_rl/pg/agents.py	2	2	1
def learn_penalty() in safe_rl/pg/agents.py	3	2	1