Some time ago I noticed that the alignment procedure uses residuals for all charged tracks (positively and negatively charged tracks together).  It was intereting to see whether residual amplitude depends on charge or not. To select different positively and negatively charged tracks I used cut on mHitsG.mCurv variable (I presume that this is curvature of the track mesured at some point) - so I am not sure whether what I call positive is actually positive as it depends on the magnetic filed direction. Need to spend some time to figure this out.

 TCut cutChargePos ="mHitsG.mCurv>0";
 TCut cutChargeNeg ="mHitsG.mCurv<0";

Fig below shows residuals for all tracks. Each pannel corresponds to one TPC sector.

Fig below shows residuals for positevely (left) and negatively (right) tracks. Each pannel corresponds to one TPC sector.