Whether this work supporting table data?

[geometrylearner]

I have test a sample by setting n_components=3 which followed error:

NotImplementedError: The current fit-tsne implementation does not support projection into dimensions other than 2 for now.

I wonder whether this algorithm improving the training on table data, if so?

[DavidMChan]

The algorithm only supports projection into two dimensions, however will take as input an arbitrary number of dimensions. For 3D projections, see #71

[geometrylearner]

Of course it supports, search in kaggle.com, the ‘rapids umap’, there is an benchmark for higher dimension. geotoppos 邮箱：geotoppos@aliyun.com Signature is customized by Netease Mail Master On 07/30/2020 12:08, David Chan wrote: The algorithm only supports projection into two dimensions, however will take as input an arbitrary number of dimensions. For 3D projections, see #71 —You are receiving this because you authored the thread.Reply to this email directly, view it on GitHub, or unsubscribe. [ { "@context": "http://schema.org", "@type": "EmailMessage", "potentialAction": { "@type": "ViewAction", "target": "#80 (comment)", "url": "#80 (comment)", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { "@type": "Organization", "name": "GitHub", "url": "https://github.com" } } ]