NormalizeFeatures

NormalizeFeatures appends scaled versions of numeric feature columns.

Inputs are columns, method, and suffix. Supported methods are zscore, robust, and minmax. Constant columns are normalized to 0.0 to keep output deterministic.

Minimal Example

from bioimageflow_core import Arguments
from bioimageflow_measurement_tools import NormalizeFeatures

normalized = NormalizeFeatures().transform(
    table,
    Arguments(columns="area,score", method="minmax"),
)

Inputs

• upstream table: feature DataFrame.

• columns: comma-separated numeric columns to normalize.

• method: zscore, robust, or minmax.

• suffix: suffix for appended normalized columns.

Outputs

The output is the upstream table plus normalized feature columns.

Dependencies and Core Libraries

pandas, NumPy-compatible numeric operations, and BioImageFlow’s DataFrameTool API.

Assumptions

Selected columns are numeric and the table fits in memory.

Expected Results

Synthetic feature tables produce deterministic scaled values and retain row order.

Failure Modes

Missing columns, unsupported methods, unreadable inputs, and invalid numeric values raise errors.