While I’m studying several machine learning models,

I saw several clustering algorithm including K-Means.

As I know K-Means use Euclidean distance for their own calculation method,
What I want is not using Euclidean distance but the value of data.

For example, Samples are distributed farily(like coordinate) and coords has their own value.
I want to find N clusters that have high value on average.

Are there any other algorithm that suits for this plot or can I just handle several parameters from K-Means algorithm to do this.

Thank you

Are there any other algorithm that suits for this plot or can I just handle several parameters from K-Means algorithm to do this.