Sklearn Truncated SVD icon

Sklearn Truncated SVD

Dimensionality reduction using truncated SVD (LSA) for sparse data

Actions2

Fit Transform

AI-generated

Summary

Fit SVD on specified feature columns and append component values and summary statistics to each workflow item.

Inputs

  • featureColumns (required) — Comma-separated list of feature column names
  • nComponents — Number of components to keep
  • algorithm — SVD solver algorithm (randomized or arpack)
  • randomState — Random seed for reproducibility
  • pythonPath — Path to Python executable with scikit-learn installed

Output shape

a list of objects, each extending the original record with fields svd_component_1, svd_component_2, … up to the chosen number of components, plus explained_variance_ratio, total_variance_explained, and svd (the full fitted SVD result)

The added fields are identical for every item; the summary statistics are computed once and attached to each output record.

Examples

Example 1: Reduce dimensionality of a text‑feature matrix

featureColumns: "feature1,feature2,feature3", nComponents: 5, algorithm: "randomized", randomState: 123, pythonPath: "python3"

Example 2: Fit with ARPACK solver and custom component count

featureColumns: "colA,colB", nComponents: 10, algorithm: "arpack", randomState: 42, pythonPath: "python3"

Example 3: Fit on a subset of records using default settings

featureColumns: "colX,colY", nComponents: 2, algorithm: "randomized", randomState: 42, pythonPath: "python3"

Discussion