[Generated by Task Expert Agent]

Performs Principal Component Analysis (PCA) on cell culture medium composition data.

This task takes a CSV file containing medium composition data and performs dimensionality reduction using PCA. It generates:

• A table of PCA scores for all samples
• A scatter plot showing PC1 vs PC2 projection
• A biplot combining scores and variable loadings

Input

• medium_csv: CSV file with medium composition data. Must contain a 'MILIEU' column identifying different medium, and numeric columns for compositional features.

Configuration

• medium_column: Name of the column containing medium identifiers (default: 'MILIEU')
• selected_medium: Optional list of specific medium to include in the analysis. If empty, all medium are included.
• decimal_separator: Decimal separator used in the CSV file (default: ',')

Outputs

• scores_table: Table containing PCA scores (principal components) for each sample
• scatter_plot: Plotly scatter plot of PC1 vs PC2 with color-coded medium groups
• biplot: Plotly biplot combining sample scores and variable loadings

Notes

• Rows with any missing (NaN) values in numeric columns are automatically removed
• Data is standardized (mean=0, std=1) before PCA
• The number of components is determined automatically based on data dimensions