dimensionality-reduction
/dimensionality-reduction
Reduce high-dimensional data to fewer components for visualization or feature engineering using PCA, t-SNE, or UMAP. Use when data has many features and needs simplification.
- Author
- chrisgve ↗
- Repository
- localdata-mcp ↗
- Licence
- Apache-2.0
- Last repository change
- 20.07.2026
Assessment and security
- Files in this package
- 1
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 0
No signals detected. Repository changed on 20.07.2026, without references to retired models or APIs.
Apache-2.0, otwarta: można używać i zmieniać, z zachowaniem licencji i informacji o zmianach.
What it does and when to use it
Reduce high-dimensional data to fewer components for visualization or feature engineering using PCA, t-SNE, or UMAP. Use when data has many features and needs simplification.
Arguments
/dimensionality-reduction <database-name>
Original author description
Reduce high-dimensional data to fewer components for visualization or feature engineering using PCA, t-SNE, or UMAP. Use when data has many features and needs simplification.
How it works
- Steps
SKILL.md preview
--- name: dimensionality-reduction description: Reduce high-dimensional data to fewer components for visualization or feature engineering using PCA, t-SNE, or UMAP. Use when data has many features and needs simplification. allowed-tools: mcp__localdata__describe_database mcp__localdata__execute_query mcp__localdata__reduce_dimensions mcp__localdata__get_data_quality_report argument-hint: "<database-name>" --- # Dimensionality Reduction Reduce data to fewer dimensions for visualization, pattern discovery, or feature engineering. ## Steps 1. **Explore features.** Call `describe_database` with the database name from `$ARGUMENTS`. Identify all numeric columns. Call `get_data_quality_report` to check for nulls and assess feature distributions. 2. **Extract data.** Call `execute_query` to select the numeric feature columns. Note the number of rows and features. High-dimensional data (10+ features) benefits most from reduction. 3. **Run PCA first.** Call `reduce_dimensions` with algorithm "pca" and 2-3 components. Review: - Explained variance ratio per component (how much information each captures) - Cumulative explained variance (target > 70% in 2-3 components for good visualization) - Component loadings (which original features contribute most to each component) 4. **Interpret PCA components.** Describe each component by its top-loading features. Name the components in domain terms when possible (e.g., "size factor" if height, weight, and volume all load heavily on PC1). 5. **Try t-SNE or UMAP for visualization.** If PCA explains less than 50% of variance in 2D (data has complex nonlinear structure), call `reduce_dimensions` with "tsne" or "umap". These methods preserve local structure better but: - Distances between distant points are not meaningful - Results depend on hyperparameters (perplexity for t-SNE, n_neighbors for UMAP) - Not suitable for downstream modeling, only for visualization 6. **Compare methods.** Assess which reduction best reveals structure: clusters, gradients, or outliers in the 2D view. PCA is interpretable; t-SNE/UMAP reveal groupings. 7. **Present results.** Provide: - Method selected and rationale - Explained variance (PCA) or stress metric - Component interpretation with feature loadings - Visual description of the 2D structure (clusters, gradients, outliers) - Recommendations: use PCA components as features in `/localdata-mcp:regression`, or use groupings visible in t-SNE as input to `/localdata-mcp:cluster-analysis`
Versions
The database has no other skills named dimensionality-reduction.
Database snapshot: 13.11.2026 · package version b82de81ad3e7 (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 95bio-data-visualization-dimensionality-reduction-plots
Produce and interpret PCA, t-SNE, UMAP, and PHATE plots for high-dimensional omics data with rigor about which method preserves what (variance, local structure, manifold, transitions), hyperparameter sensitivity, and the well-documented limits of 2D embeddings. Covers PCA… - 100pca-dimensionality-reduction
Use when performing PCA principal component dimensionality reduction on tabular numeric data. Supports command-line parameter input, automatic numeric feature selection, parameter validation, result directory creation, and CSV or TXT format result export. - 100umap-learn
Use when reducing high-dimensional data to 2D/3D for visualization, preprocessing embeddings before clustering with HDBSCAN, or training parametric dimensionality-reduction models. - 95fiftyone-embeddings-visualization
Visualizes datasets in 2D using embeddings with UMAP or t-SNE dimensionality reduction. Use when exploring dataset structure, finding clusters, identifying outliers, or understanding data distribution. - 100alterlab-umap
Nonlinear dimensionality reduction with UMAP — fast manifold learning for 2D/3D visualization, clustering preprocessing (e.g., HDBSCAN), and supervised or parametric UMAP. Use when projecting high-dimensional data to low dimensions for visualization, embedding generation, or as… - 100scanpy
Standard single-cell RNA-seq analysis pipeline. Use for QC, normalization, dimensionality reduction (PCA/UMAP/t-SNE), clustering, differential expression, and visualization. Best for exploratory scRNA-seq analysis with established workflows. For deep learning models use…