Different embeddings for identical data

[Baschdl]

Why does transform return different embeddings for the same data?

model = UMAP_(transpose(data), 2; n_neighbors=15, min_dist=0.1)
UMAP.transform(model, transpose(data[1:20,:]))

results in

2×20 Matrix{Float32}:
-7.9815   8.59674  -7.18658  7.90244  …  -2.41209   2.69043   4.15093
-2.66961  0.89672  -3.72272  6.20439      7.94941  -7.61004  -8.79426

Transforming the same data a second time results in a different embedding:

UMAP.transform(model, transpose(data[1:20,:]))

results in

2×20 Matrix{Float32}:
 -7.99711  8.49432  -7.27412  7.844    …  -2.32314   2.682     4.33971
 -2.75653  1.0201   -3.63024  5.85082      8.06509  -7.60372  -8.87176

Do I have to fix a certain random seed to get identical embeddings?