Kali: The framework for fine calibration of LHCb Electromagnetic Calorimeter

The precise calibration of the electromagnetic calorimeter (ECAL) of the LHCb experiment is an essential task for the fulfilment of the LHCb physics program. The goal of the calibration of the ECAL is to intercalibrate the 6016 cells at a alevel better than 2% and provide an overall calibration below the 1% level. The final step of this task is performed with two calibration methods using the real data from the experimental setup. Both of them require processing of $\mathcal{O}(10^8)$ events which must be selected, reconstructed and analyzed. The analysis is very CPU-consuming, since one method performs an adaptive multi-pass fitting of $\mathcal{O}(2\times( 10^5)$ histograms and the other one the minimization of $\mathcal{O}(1.5\times10^5)$ data-based functions.In this document we present the Kali framework developed within the LHCb software framework, which implements these two final calibration methods. It is integrated with Grid and makes use of parallelism tools, such as python parallel processing modules, to provide an efficient way, both time and disk wise, for the final ECAL calibration.The results of the fine calibration with the very first data collected by the LHCb experiment will also be presented. With the use of the Kali framework it took only 2 days of processing and allowed to achieve a miscalibration level of 2-2.5 % for the different ECAL areas.