transform_outliers() [2/2]

template<OperationReadyData DataType, typename TransformFunc >
requires std::invocable<TransformFunc, double>

DataType MayaFlux::Yantra::transform_outliers	(	DataType &	input,
		double	std_dev_threshold,
		TransformFunc	transform_func,
		std::vector< std::vector< double > > &	working_buffer
	)

Statistical outlier transformation using existing StatisticalAnalyzer (OUT-OF-PLACE)

Template Parameters

DataType	OperationReadyData type
TransformFunc	Function type for transformation

Parameters

input	Input data - will NOT be modified
std_dev_threshold	Standard deviation threshold for outliers
transform_func	Transformation function
working_buffer	Buffer for operations (will be resized if needed)

Returns

Transformed data

Definition at line 276 of file MatrixHelper.hpp.

{
    auto [target_data, structure_info] = OperationHelper::setup_operation_buffer(input, working_buffer);
 
    auto stat_analyzer = std::make_shared<StandardStatisticalAnalyzer>();
    auto stats = stat_analyzer->analyze_statistics(input);
 
    if (stats.channel_statistics.empty()) {
        throw std::runtime_error("No channel statistics available for outlier detection");
    }
    const auto& first_channel_stats = stats.channel_statistics[0];
    double threshold_low = first_channel_stats.mean_stat - std_dev_threshold * first_channel_stats.stat_std_dev;
    double threshold_high = first_channel_stats.mean_stat + std_dev_threshold * first_channel_stats.stat_std_dev;
 
    for (auto& channel_span : target_data) {
        std::ranges::transform(channel_span, channel_span.begin(),
            [&](double x) {
                return (x < threshold_low || x > threshold_high) ? transform_func(x) : x;
            });
    }
 
    return OperationHelper::reconstruct_from_double<DataType>(working_buffer, structure_info);
}

References MayaFlux::Yantra::OperationHelper::setup_operation_buffer().

Here is the call graph for this function: