amazon-research / meta-learning-the-difference
Conditional Complexity

The distribution of complexity of units (measured with McCabe index).

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 361 units with 4,720 lines of code in units (72.8% of code).
    • 0 very complex units (0 lines of code)
    • 2 complex units (154 lines of code)
    • 24 medium complex units (1,061 lines of code)
    • 50 simple units (1,169 lines of code)
    • 285 very simple units (2,336 lines of code)
0% | 3% | 22% | 24% | 49%
Legend:
51+
26-50
11-25
6-10
1-5
Alternative Visuals
Conditional Complexity per Extension
51+
26-50
11-25
6-10
1-5
py0% | 3% | 22% | 24% | 49%
perl0% | 0% | 0% | 0% | 100%
Conditional Complexity per Logical Component
primary logical decomposition
51+
26-50
11-25
6-10
1-5
low_rank_comparisons/src0% | 6% | 21% | 18% | 53%
abstractive_summarization/src/others0% | 10% | 11% | 8% | 69%
dialogue_personalization/utils0% | 0% | 29% | 32% | 38%
abstractive_summarization/src0% | 0% | 40% | 17% | 42%
dialogue_personalization/model0% | 0% | 10% | 37% | 51%
low_rank_comparisons/eval0% | 0% | 15% | 36% | 48%
dialogue_personalization0% | 0% | 0% | 58% | 41%
Most Complex Units
Top 20 most complex units
Unit# linesMcCabe index# params
def build_optim()
in abstractive_summarization/src/others/optimizer.py
 58 42 4
def forward()
in low_rank_comparisons/src/model.py
 96 33 5
def beam_search_sample()
in dialogue_personalization/utils/beam_ptr.py
 72 23 8
def forward()
in low_rank_comparisons/src/model.py
 55 18 10
def convert_examples_to_features()
in dialogue_personalization/utils/load_bert.py
 43 18 4
def forward()
in low_rank_comparisons/src/model.py
 46 17 6
def evaluate()
in dialogue_personalization/model/common_layer.py
 39 16 8
def translate_batch()
in dialogue_personalization/utils/beam_omt.py
 80 16 2
def train()
in abstractive_summarization/src/dapt_pretraining.py
 47 16 1
def train()
in abstractive_summarization/src/trainer.py
 55 16 7
def train()
in abstractive_summarization/src/tapt_pretraining.py
 47 16 1
def get_balanced_loader()
in dialogue_personalization/utils/data_reader.py
 42 15 6
def parse()
in low_rank_comparisons/eval/eval.py
 29 14 4
def beam()
in low_rank_comparisons/src/gpt2_beam.py
 85 14 3
def train_validate()
in low_rank_comparisons/src/gpt2_ft.py
 57 14 8
def train()
in abstractive_summarization/src/sdpt_pretraining.py
 49 14 6
def step()
in low_rank_comparisons/src/optimizer.py
 31 13 2
def preprocess()
in dialogue_personalization/utils/data_reader.py
 31 13 2
def get_data_loader()
in dialogue_personalization/utils/data_reader.py
 36 13 5
def set_parameters()
in abstractive_summarization/src/others/optimizer.py
 30 12 2