source: fact/tools/rootmacros/SpikeRemoval.C@ 12572

// compute the mean of the left and right neighbors of a channel
vector<float> * computeN1mean( vector<float> &src)
{
        vector<float> * dest = new vector<float>;

        dest->push_back(src[1]);

        for(unsigned int i = 1; i < src.size() - 1; i++){
/*        if (i == 0){ // use right sample es mean
                        N1mean[i] = Ameas[i+1];
                }
                else if ( i == Samples-1 ){ //use left sample as mean
                        N1mean[i] = Ameas[i-1];
                }
                else{
                        N1mean[i] = ( Ameas[i-1] + Ameas[i+1] ) / 2.;
                }
*/
                dest->push_back( ( src[i-1] + src[i+1] ) / 2. );
        }

        dest->push_back(src[src.size() - 1]);
        return dest;
} // end of computeN1mean computation

void removeSpikes(
        vector<float> &src,
        vector<float> &dest,
        const float CandidateTHR,
        const float nextDiffTHR,
        const float nextNextDiffTHR
){
        vector<float> * NextNeighborMean = computeN1mean(src);

        if (src.size() != NextNeighborMean->size())
                //TODO .. if verbositylevel .. say something, use the return code to tell the caller..
        dest.clear();
        dest = src;

        float diff, nextDiff, nextNextDiff;

// find the spike and replace it by mean value of neighbors
        for (unsigned int i = 0; i < src.size(); i++) {

                diff = src[i] - (*NextNeighborMean)[i];

                if ( diff < CandidateTHR ){ // a spike candidate
                        // check consistency with a single channel spike

                        if ( src[i+2] - ( src[i] + src[i+3] )/2. > 10. )
                        {
                                dest[i+1] = ( src[i] + src[i+3] )/2.;
                                dest[i+2] = ( src[i] + src[i+3] )/2.;
                                i = i + 3;
                        }
                        else
                        {
                                nextDiff = src[i+1] - (*NextNeighborMean)[i+1];
                                nextNextDiff = src[i+2] - (*NextNeighborMean)[i+2];

                                if ( ( nextDiff > nextDiffTHR * diff ) && ( nextNextDiff < nextNextDiffTHR ) ){
                                        dest[i+1] = (*NextNeighborMean)[i+1];
                                        (*NextNeighborMean)[i+2] = (src[i+1] - src[i+3] / 2.);
                                        i = i + 2;//do not care about the next sample it was the spike
                                }
                                // treatment for the end of the pipeline must be added !!!
                        }
                }
        } // end of spike search and correction
        delete NextNeighborMean;
        return;
}