This section from the older manual repeats in some parts what have been said before, but Ive decided to leave it due to some useful details mentioned.
Data analysis in NEMO-3 experiment is performed in several steps.
First step is to reconstruct rawdata and MC with nemor reconstruction program. It applies energy and time calibration for calorimeter data, Geiger calibrations for tracking data, reproduce detector configuration for correct efficiency calculation with MC. Also it performs event tracking. At UCL NEMO CVS stored version of nemor program designed to use with combination of all other rootana package. It saves calorimeter, geiger and tracker information blocks are into h10 ROOT tree. Also is fills variables run, date and time for REAL and MTCA files which is usefully for normalisation (see Normalisation section). There are some PERL scripts to setup nemor environment, run reconstruction and convert files to ROOT with h2root tool.
Second step is rough event preselecting. Amount of data after reconstruction is huge and difficult to handle. To make efficient analysis one need to reduce it. Wise compromise should be done at this step. If final analysis and histogram plotting takes more then several minutes it will slow down your actual work and will be waste of time for waiting the results in front of the monitor. From the other hand, too rigorous cuts restrict flexibility of final analysis. You may want to look at smth. you cut off earlier. Useful program skeleton, to perform reduction, so-called slimming, is in rootana/slim2 directory. See slim2 section for further explanation how to use it. It is good style to produce one slim file for each MC component and for data sample.
Third step is to plot histograms, using slimmed files, and make final analysis. A skeleton of analysis program rootana/ana10 was developed in order to help end users to manage copies of similar histograms, make fit and fancy drawings. See section ana10 for further details.