RQ4 : In terms of the impact of code clones,what is the difference between projects using different frameworks?

Glossary

Group1 : contains projects that only utilize Pytorch framework to support deep learning;
Group2 : contains projects that only utilize Tensorflow framework to support deep learning;
Group3 : contains projects that can both Pytorch and Tensorflow framework to support deep learning;
Group4 : if we couldn’t determine which kind of framework was used in a project based on its description and contents,we then categorized this project into this group;
Pct : the percentage of cloned LOC to the LOC of a project;
PCC : the percentage of the co-changed clones to all clones;
PCC1, PCC2, PCC3 : the percentage of the number of co-changed code cloens to all clones for each type of clones;
AvgPct, AvgPCC, AvgPCC1, AvgPCC2, AvgPCC3 : the corresponding average;

Group1	Group2	Group3	Group4
tianshou	tflearn	deepchem	imgaug
ignite	TensorLayer	horovod	nni
DIG	pyod	transformers	chainer
pytorchvideo	texar	clearml
torch-points3d	sonnet	DeepPavlov
ludwig	keras	PySyft
catalyst	addons	spaCy
allennlp	tfx	ray
TTS	autokeras	Hub
raster-vision	stellargraph
MONAI	tensor2tensor
pytorch-lightning	luminoth
OpenNMT-py	coach
torchio	DeepLabCut
	faceswap
	deepvariant
	DeepFaceLab
	GPflow
	tensorpack

	AvgPct	AvgPCC	AvgPCCT1	AvgPCCT2	AvgPCCT3
Group 1	15.7%	18.8%	9.1%	19.3%	27.6%
Group 2	15.4%	15.5%	12.8%	12.7%	18.8%
Group 3	13.2%	6.2%	1.7%	2.4%	9.6%

Subject	Pct	PCC	PCCT1C	PCCT2C	PCCT3C
tianshou	51.10%	77.9%	0	0	78.2%
catalyst	13.50%	39.5%	0	75.0%	35.5%
ludwig	15.00%	38.7%	5.6%	31.4%	40.5%
MONAI	7.00%	32.3%	54.5%	6.3%	33.6%
DIG	20.60%	23.7%	6.7%	28.6%	26.8%
torch-points3d	16.00%	10.2%	15.4%	14.3%	9.3%
allennlp	12.20%	7.5%	0	0	9.0%
TTS	10.30%	7.0%	0	3.6%	9.3%
ignite	38.90%	7.0%	0	14.8%	6.0%
raster-vision	8.20%	0
pytorch-lightning	4.00%	0
OpenNMT-py	3.40%	0
torchio	3.20%	0
pytorchvideo	17.00%
Avg	15.7%	18.8%	9.1%	19.3%	27.6%

Subject	Pct	PCC	PCCT1C	PCCT2C	PCCT3C
tflearn	35.40%	62.1%	0	50.0%	64.9%
deepvariant	4.50%	31.3%	33.3%	33.3%	29.7%
pyod	25.70%	24.7%	15.4%	29.5%	23.2%
stellargraph	15.10%	20.6%	0	22.2%	20.4%
autokeras	16.70%	18.8%	100.0%	0	13.2%
coach	10.80%	17.9%	0	0	22.7%
texar	23.10%	17.5%	6.7%	16.7%	18.6%
DeepLabCut	9.50%	16.7%	0	2.4%	28.1%
keras	21.20%	16.0%	29.4%	10.2%	16.7%
TensorLayer	29.70%	14.2%	1.3%	19.4%	15.7%
tensorpack	0.80%	12.5%	0	0	16.7%
luminoth	11.90%	4.5%	0	0	5.3%
tfx	20.00%	3.7%	0	5.6%	3.6%
addons	20.10%	2.6%	0	0	3.2%
tensor2tensor	13.40%	0.7%	6.3%	1.5%	0.5%
sonnet	22.40%	0
GPflow	3.40%	0
faceswap	5.60%
DeepFaceLab	3.80%
Avg	15.4%	15.5%	12.8%	12.7%	18.8%

Subject	Pct	PCC	PCCT1C	PCCT2C	PCCT3C
deepchem	28.20%	4.5%	0	2.8%	4.9%
clearml	10.50%	17.3%	0	4.0%	26.8%
Hub	5.20%	16.7%	0	0	17.9%
transformers	21.60%	6.6%	12.0%	6.3%	5.8%
ray	6.40%	5.9%	0	4.0%	6.7%
horovod	22.50%	2.8%	0	0	3.2%
DeepPavlov	8.70%	2.0%	0	0	2.0%
PySyft	8.40%	0
spaCy	6.90%	0
Avg	13.2%	6.2%	1.7%	2.4%	9.6%

Subject	Pct	PCC	PCCT1C	PCCT2C	PCCT3C
nni	23.80%	3.1%	2.5%	17.3%	2.2%
imgaug	28.80%	15.9%	40.0%	12.6%	16.1%
chainer	19.20%	2.4%	0	6.4%	2.1%
Avg	23.9%	7.1%	14.2%	12.1%	6.8%