与聚类混淆

Question

I am getting so confused with clustering in data science process. We know that the process of grouping similar points in a 2D space is based on this formula:

distance = sqrt( (x2-x1)^2 + (y2-y1)^2 )

But in introducing inputs to the sklearn we just feed the x-axis values :( what happened to the y-axis values?

for example we have the following data base:

index    x     y
------------------
  0      5     8
  1      6     9
  2      7     10

and we introduce x to the KMeans

from sklearn.cluster import KMeans
kmeans = KMeans(2)
kmeans.fit(df["x"])

How can it calculate distance without having y values?

Answer 1

KMeans clustering can be done in any number of dimensions. As you said, the distance can be calculated using the Euclidean distance . This distance can be calculated for any number of dimension. You passed one array, so in this case it's just one dimension, so the formula would simplify to:

distance = sqrt((x2-x1)^2)

Which is really just the absolute value of (x2-x1)