alibaba / TinyNeuralNetwork - 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 2,186 units with 35,903 lines of code in units (80.0% of code).

7 very complex units (2,715 lines of code)
14 complex units (1,657 lines of code)
116 medium complex units (5,476 lines of code)
134 simple units (3,120 lines of code)
1,915 very simple units (22,935 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 rewrite_quantize_graph() in tinynn/graph/quantization/quantizer.py	1175	264	2
def process_op() in tinynn/converter/operators/op_version.py	283	117	2
def elementwise_op_reshape_passthrough_pass() in tinynn/converter/operators/optimize.py	374	104	1
def elementwise_reshape_transpose_passthrough_pass() in tinynn/converter/operators/optimize.py	233	86	1
def calc_prune_idx() in tinynn/graph/modifier.py	200	82	4
def elementwise_op_transpose_passthrough_pass() in tinynn/converter/operators/optimize.py	282	78	2
def elementwise_op_quantize_passthrough_pass() in tinynn/converter/operators/optimize.py	168	55	1
def output_transpose_pass() in tinynn/converter/operators/optimize.py	174	47	1
def disable_requantization_for_cat_pass() in tinynn/graph/quantization/quantizer.py	123	45	2
def modify_output() in tinynn/graph/modifier.py	126	41	2
def transform() in tinynn/converter/operators/tflite/transformable.py	181	39	3
def prepare_onnx_export() in tinynn/graph/quantization/quantizer.py	148	37	2
def transform() in tinynn/converter/operators/tflite/transformable.py	190	37	3
def register_mask() in tinynn/graph/modifier.py	114	34	4
def hook_modules() in tinynn/graph/tracer.py	96	32	1
def parse_args() in tinynn/graph/tracer.py	110	30	3
def prepare_qconfig() in tinynn/graph/quantization/quantizer.py	96	27	3
def rewrite_quantize_graph() in tinynn/graph/quantization/quantizer.py	85	27	2
def __gen_forward_code() in tinynn/graph/tracer.py	57	27	2
def topological_sort() in tinynn/converter/operators/graph.py	57	27	1
def parse() in tinynn/converter/operators/torch/aten.py	100	26	5
def prepare_qconfig() in tinynn/graph/quantization/quantizer.py	121	25	3
def parse() in tinynn/converter/operators/torch/aten.py	113	25	5
def error_analysis() in tinynn/graph/quantization/quantizer.py	89	24	4
def __tag_nodes() in tinynn/graph/tracer.py	65	24	1
def quantize_input_output_type_pass() in tinynn/converter/operators/optimize.py	85	24	1
def is_ending_with_noop_edge() in tinynn/converter/operators/optimize.py	51	24	3
def gen_module_constrctor_line() in tinynn/graph/tracer.py	78	23	2
def handle_reduce() in tinynn/converter/operators/torch/base.py	65	22	7
def parse() in tinynn/converter/operators/torch/aten.py	100	22	5
def parse_common() in tinynn/converter/operators/torch/aten.py	88	22	6
def parse_tflite() in contrib/tflite2tf/tflite2tf.py	95	21	1
def __init__() in tinynn/graph/quantization/quantizer.py	48	20	5
def reshape_mapping() in tinynn/converter/operators/optimize.py	60	20	2
def __init__() in tinynn/prune/oneshot_pruner.py	37	19	4
def rebuild() in tinynn/graph/modifier.py	49	19	1
def modify_input() in tinynn/graph/modifier.py	70	19	2
def graph_error_analysis() in tinynn/util/quantization_analysis_util.py	63	18	3
def rewrite_quantize_graph_for_tensorrt() in tinynn/graph/quantization/quantizer.py	46	18	2
def prev_node_unique_name() in tinynn/graph/tracer.py	38	18	3
def add_forward_node() in tinynn/graph/tracer.py	73	18	3
def insert_after() in tinynn/graph/tracer.py	48	18	4
def parse_common() in tinynn/converter/operators/torch/quantized.py	68	18	4
def layer_error_analysis() in tinynn/util/quantization_analysis_util.py	71	17	4
def update_weight_metric() in tinynn/graph/modifier.py	41	17	4
def dim_change_forward() in tinynn/graph/modifier.py	38	17	6
def prepare_torch_overrides_funcs() in tinynn/graph/tracer.py	37	17	1
def fuse_requantize() in tinynn/converter/operators/optimize.py	59	17	1
def group_conv_rewrite_pass() in tinynn/converter/operators/optimize.py	72	17	1
def group_deconv_rewrite_pass() in tinynn/converter/operators/optimize.py	73	17	1