High Dimensional Data Visualization with Clustergrammer2 |SciPy 2020| Nicolas Fernandez

preview_player
Показать описание
Visualizing complex, high-dimensional data is a key step in data analysis and is traditionally approached using dimensionality reduction techniques (e.g. t-SNE, UMAP). We propose an alternative and complementary approach using heatmaps and the interactive Python Widget and WebGL visualization library Clustergrammer2. Clustergrammer2 enables users to interactively explore and intuitively understand high-dimensional data structures consisting of millions of datapoints with ease. We demonstrate how Clustergrammer2 can be used on single-cell gene expression data as well as novel spatial transcriptomics data to identify cell types/states as well as derive novel phenotypic signatures from data driven clustering.

*****************
Рекомендации по теме
Комментарии
Автор

Interesting, definitely a good addition to the traditional t-sne / umap approach to visualization.

kayae
Автор

Hi Nick, great talk! could you send the link to the articles analysis related to COVID-19?

horaciovargasguzman