ExpressionExperimentBatchCorrectionService (Gemma 1.32.1 API)

All Known Implementing Classes:

ExpressionExperimentBatchCorrectionServiceImpl
```
public interface ExpressionExperimentBatchCorrectionService
```
Author:

Paul

Method Summary

All Methods Instance Methods Abstract Methods
Modifier and Type	Method	Description
`boolean`	`checkCorrectability(ExpressionExperiment ee)`	Is there a Batch factor provided?
`ExpressionDataDoubleMatrix`	`comBat(ExpressionExperiment ee)`	Run ComBat using default settings (parametric)
`ExpressionDataDoubleMatrix`	`comBat(ExpressionExperiment ee, ExpressionDataDoubleMatrix mat)`	Run ComBat with a specific data matrix.
`ExperimentalFactor`	`getBatchFactor(ExpressionExperiment ee)`	For convenience of some testing classes

- Method Detail
  - checkCorrectability
```
boolean checkCorrectability(ExpressionExperiment ee)
```
    Is there a Batch factor provided? Is there a confound problem? Do we have at least two samples per batch? This will return true even if there is evidence the data has been batch-corrected before; we assume the caller wants to redo it based on the raw data
    
    Parameters:
    
    ee - the experiment
    
    Returns:
    
    whether it is correctable
  - comBat
```
@Nullable
ExpressionDataDoubleMatrix comBat(ExpressionExperiment ee)
```
    Run ComBat using default settings (parametric)
    The matrix is constructed from the experiment's processed vectors.
  - comBat
```
@Nullable
ExpressionDataDoubleMatrix comBat(ExpressionExperiment ee,
                                  ExpressionDataDoubleMatrix mat)
```
    Run ComBat with a specific data matrix.
    
    Parameters:
    
    mat - the matrix
    
    Returns:
    
    batch corrected matrix, or null if there's no batching factor
  - getBatchFactor
```
@Nullable
ExperimentalFactor getBatchFactor(ExpressionExperiment ee)
```
    For convenience of some testing classes
    
    Parameters:
    
    ee - the experiment to get the batch factor for
    
    Returns:
    
    the batch factor of the experiment, or null, if experiment has no batch factor