openai / vdvae - 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 83 units with 848 lines of code in units (84.5% of code).

0 very complex units (0 lines of code)
0 complex units (0 lines of code)
5 medium complex units (149 lines of code)
4 simple units (120 lines of code)
74 very simple units (579 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 train_loop() in train.py	32	16	7
def accumulate_stats() in train_helpers.py	22	13	2
def discretized_mix_logistic_loss() in vae_helpers.py	48	12	3
def restore_log() in train_helpers.py	9	12	3
def logger() in utils.py	38	11	1
def training_step() in train.py	18	10	7
def set_up_data() in data.py	74	10	1
def parse_layer_string() in vae.py	17	8	1
def parse_args_and_update_hparams() in hps.py	11	6	3
def evaluate() in train.py	10	5	4
def first_rank_first() in train_helpers.py	8	5	2
def restore_params() in train_helpers.py	12	5	6
def load_vaes() in train_helpers.py	22	5	2
def cifar10() in data.py	19	5	2
def maybe_download() in utils.py	16	4	2
def write_images() in train.py	13	4	6
def build() in vae.py	13	4	1
def build() in vae.py	16	4	1
def forward_uncond() in vae.py	12	4	4
def get_conv() in vae_helpers.py	7	3	9
def allreduce() in utils.py	4	3	2
def get_cpu_stats_over_ranks() in utils.py	4	3	1
def tile_images() in utils.py	15	3	4
def get_sample_for_visualization() in train.py	6	3	4
def load_opt() in train_helpers.py	12	3	3
def __init__() in vae.py	8	3	8
def forward() in vae.py	9	3	2
def get_width_settings() in vae.py	8	3	2
def forward() in vae.py	11	3	2
def sample_uncond() in vae.py	12	3	4
def get_inputs() in vae.py	9	3	3
def forward() in vae.py	11	3	4
def forward_uncond() in vae.py	13	3	4
def forward() in vae.py	8	3	3
def forward_manual_latents() in vae.py	8	3	4
def flatten() in data.py	2	3	1
def sample_from_discretized_mix_logistic() in vae_helpers.py	18	2	2
def __getattr__() in utils.py	5	2	2
def num_nodes() in utils.py	5	2	0
def gpus_per_node() in utils.py	5	2	0
def run_test_eval() in train.py	7	2	5
def main() in train.py	8	2	0
def update_ema() in train_helpers.py	4	2	3
def linear_warmup() in train_helpers.py	4	2	1
def distributed_maybe_download() in train_helpers.py	7	2	3
def set_up_hyperparams() in train_helpers.py	15	2	1
def __getattr__() in hps.py	5	2	2
def forward() in vae.py	11	2	3
def unpickle_cifar10() in data.py	6	2	1
def ffhq1024() in data.py	2	2	1