來(lái)自于公眾號(hào)：DLCV

1. tensorboard和tensorboardX被開發(fā)用來(lái)支持chainer, mxnet, numpy。https://github.com/lanpa/tensorboardX

tensorboard

2. visdom，常用于pytorch數(shù)據(jù)可視化，也支持numpy。https://github.com/facebookresearch/visdom

3. seaborn：一款基于matplotlib擁有更高API的工具，主要處理低維數(shù)據(jù)。https://github.com/mwaskom/seaborn

4. holoviews：與season差不多。https://github.com/ioam/holoviews

5. missingno：一款缺失數(shù)據(jù)可視化工具，適合分析數(shù)據(jù)集的完整性。https://github.com/ResidentMario/missingno