plotting on an xy plot
lketdegh opened this issue · comments
Hi @LTLA ,
I don't understand much about how to use it.
UMAP is work for reduct dimemsion but how do I draw the in xy plot via result?
Is there a reason you're using a C++ library? If you just want to create and plot UMAPs, I'd suggest you start with either the original Python package or the uwot R package, which operate in frameworks that have more plotting functionality. If you must use C++, I suppose gnuplot could work.
@LTLA Thank a lot!
My data is generated by C++ at all that's the reason I trying to using C++ library. And I have no idea how to realize the output data because even the value of outDim is 2 but output 1 dim vector only. I don't get how could I to generate the 2 dimension result via n dimension data? Thank you for your help and patience.
Or in other words, I got a graph with multiple clusters via the Python package and I hope I can reproduce it using C++. Is there any method to make data to 2 dimension result and I could plotting it myself.
Hello @LTLA ,
Thank you for help about solve the problem above. The plot made with several cluster already like I want. Is there any way to know each cluster belongs to what data is it like as the photo below from Python package ?
Hello @LTLA , Thank you for help about solve the problem above. The plot made with several cluster already like I want. Is there any way to know each cluster belongs to what data is it like as the photo below from Python package ?
UMAP doesn't do clustering. The iris
dataset you are referencing in the image above is already labeled via its species (setosa, versicolor, virginica) before UMAP was run on it. For clustering HDBSCAN is often used with UMAP, but trying to do all of this in C++ is going to require you to write a lot of code on your own. Otherwise, if you can switch to Python or R for further data analysis and visualization I would recommend it.
@jlmelville @LTLA
Really appreciate your help in resolving the problem and it's useful for me.