Direct comparison
UMAP vs t-SNE for Dimensionality Reduction
UMAP vs t-SNE for visualizing high-dimensional data: what each preserves, the cluster-distance pitfall, speed at scale, and which to pick.
Written and maintained by CASRAI Editorial Board
Last updated
Ask CASRAI · free to try
Ask about UMAP vs t-SNE for Dimensionality Reduction
Ask your first 2 questions free below. Subscribers get 150 a day for $29 a month.
An AI assistant specialized in research administration. It cites the sources behind every answer, labels web answers and says when it can't answer.
Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.
Works on this site and inside Claude, Cursor and the AI tools you already use.
Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays free to read, with no account and no card.
How do UMAP, t-SNE compare side by side?
The table below compares UMAP, t-SNE across 12 procurement-relevant dimensions, from full name through axis interpretability.
Side-by-side comparison
| Dimension | UMAP | t-SNE |
|---|---|---|
| Full name | Uniform Manifold Approximation and Projection | t-distributed Stochastic Neighbor Embedding |
| Introduced | McInnes, Healy & Melville, arXiv:1802.03426 (2018, rev. 2020) | van der Maaten & Hinton, JMLR 9(86):2579-2605 (2008) |
| Core mechanism | Builds a fuzzy topological (simplicial-set) graph of nearest neighbors, then optimizes a low-dimensional layout to match it | Converts pairwise similarities to probabilities in high-D and low-D (heavy-tailed t-distribution), minimizes KL divergence between the two |
| Local structure | Strong -- visually competitive with t-SNE per the original paper | Strong -- this is what t-SNE is specifically optimized for; produces tight, well-separated islands |
| Global structure | Authors' paper states it "arguably preserves more of the global structure" than t-SNE | Weaker -- optimized for local neighborhoods; relative positions of distant clusters are not reliable |
| Between-cluster distances | Not a calibrated distance -- do not read gap size as a quantitative difference | Documented limitation: cluster size and inter-cluster distance in the plot are not informative measurements |
| Key hyperparameter | n_neighbors (local neighborhood size) and min_dist (how tightly points pack) | Perplexity (roughly, effective neighbor count), typically 5-50 |
| Speed at scale | Reported by its authors as consistently faster, with near-linear scaling and no penalty for embedding into more than 2-3 dimensions | Slower on large datasets; classic implementation scales less favorably, motivating Barnes-Hut and FIt-SNE approximations |
| Reproducibility across runs | Layout can shift with random seed and hyperparameters; not guaranteed identical run-to-run | Same caveat -- known to be sensitive to random initialization and perplexity choice |
| Feeding a downstream algorithm | Its neighbor graph is commonly reused directly by downstream tools (e.g., graph-based clustering) | Typically used only as a terminal visualization, not as input to further steps |
| Typical default in single-cell pipelines (2026) | Most common default visualization step in current single-cell toolkits | Still widely used, especially where tight local separation is the priority |
| Axis interpretability | Unitless, non-interpretable individually -- no loading/variance-explained meaning like a PCA biplot | Same -- axes carry no independent quantitative meaning |
Common questions
Common questions about UMAP vs t-SNE
Does UMAP preserve global structure better than t-SNE?
+
That is the claim UMAP's original authors make in the introducing paper -- that it "arguably preserves more of the global structure" while remaining visually competitive locally. Treat it as the algorithm's design goal and the authors' reported result, not an independently settled distance-accuracy guarantee.
Can I trust the distance between two clusters in a UMAP or t-SNE plot?
+
No, not as a quantitative measurement. Both algorithms optimize local neighborhood preservation; the gap between clusters and the apparent size of a cluster are artifacts of the optimization and hyperparameters, not calibrated distances or sample counts. Verify any global pattern against the original high-dimensional data before reporting it as a finding.
Is UMAP always faster than t-SNE?
+
It is reported as consistently faster in the original comparison, and that holds up in common practice, particularly as dataset size grows into the tens of thousands of points and beyond. On small datasets the difference is usually negligible.
Should I run PCA before UMAP or t-SNE?
+
Yes -- this is standard practice, especially for high-dimensional data like single-cell expression matrices. Reducing to a moderate number of principal components first cuts noise and computation before either nonlinear method runs its neighbor search.
Do perplexity (t-SNE) and n_neighbors (UMAP) mean the same thing?
+
They play an analogous role -- both roughly set the effective local neighborhood size -- but they are not mathematically identical, and default values differ by implementation. Changing either meaningfully changes how tight or diffuse the resulting embedding looks.
