Class RowMissingValueFilter
- java.lang.Object
-
- ubic.gemma.core.analysis.preprocess.filter.RowMissingValueFilter
-
public class RowMissingValueFilter extends Object
Filter out rows that have "too many" missing values.- Author:
- pavlidis
-
-
Constructor Summary
Constructors Constructor Description RowMissingValueFilter()
-
Method Summary
All Methods Instance Methods Concrete Methods Modifier and Type Method Description ExpressionDataDoubleMatrix
filter(ExpressionDataDoubleMatrix data)
void
setAbsentPresentCalls(ExpressionDataBooleanMatrix absentPresentCalls)
Supply a separate matrix of booleans.void
setMaxFractionRemoved(double f)
Set the maximum fraction of rows which will be removed from the data set.void
setMinPresentCount(int m)
Set the minimum number of values that must be present in each row.void
setMinPresentFraction(double k)
-
-
-
Method Detail
-
filter
public ExpressionDataDoubleMatrix filter(ExpressionDataDoubleMatrix data)
-
setAbsentPresentCalls
public void setAbsentPresentCalls(ExpressionDataBooleanMatrix absentPresentCalls)
Supply a separate matrix of booleans. This is not necessary if the input matrix is already 'masked' for missing values.- Parameters:
absentPresentCalls
- new value
-
setMaxFractionRemoved
public void setMaxFractionRemoved(double f)
Set the maximum fraction of rows which will be removed from the data set. The default value is 0.3 Set it to 1.0 to remove this restriction.- Parameters:
f
- double
-
setMinPresentCount
public void setMinPresentCount(int m)
Set the minimum number of values that must be present in each row. The default value is 5. This is always overridden by a hard-coded value (currently 2) that must be present for a row to be kept; but this value is in turn overridden by the maxFractionRemoved.- Parameters:
m
- int
-
setMinPresentFraction
public void setMinPresentFraction(double k)
- Parameters:
k
- double the fraction of values to be removed.
-
-