WS Calibration (PSB) Student Meeting 11/08/2014 Carolina Bianchini BE-BI-BL
11/08/2014 Carolina Bianchini 2 Outline Program timing Conclusions Data Structure
11/08/2014 Carolina Bianchini 3 Program timing 1. Select the Calibration Folder 2.Find the peaks value and calculate the average value between them (at each position by the relative.xml file). 3.Data Processing: choose the best polynomial fit 4.Generate the Calibration Table and write the Calibration Table in.clt file
11/08/2014 Carolina Bianchini 4 Program timing READING first structure TIME E Tempo di lettura file xml READING second structure TIME E-02 SMOOTHING TIME E E E-01 SMOOTHING TIME 2STEP E-03 double windows fit time E-02 SORTING MATRIX TIME E+02 PS time256 s PSB168 s
11/08/2014 Carolina Bianchini 5 Data Structure Serie data dir speed coef (dict) meanFit (dict) rmsFit (dict) fit_goodness (dict) dist (dict) dMean (dict) dRMS (dict) Calibration_Program (data) pos center1 center2 avg type speed
11/08/2014 Carolina Bianchini 6 Data Structure This is the class object Structures. The program makes a dictionary of data type Calibration_Program for each position: inRes = Calibration_Program(pos * parameter, center1In, center2In, AvgCenterIn, 'in', speed) homeRes = Calibration_Program(pos * parameter, center1Home, center2Home, AvgCenterHome, 'home', speed) Then collect these data in different dictionary for speed (10/15) and movement (In/Out): dict10IN[inRes.pos].append(inRes) dict10HOME[homeRes.pos].append(homeRes) dict15IN[inRes.pos].append(inRes) dict15HOME[homeRes.pos].append(homeRes) And insert all in a Series Class: s10In = Serie(dict10IN, 10, 'in') s15In = Serie(dict15IN, 15, 'in') s10Home = Serie(dict10HOME, 10, 'home') s15Home = Serie(dict15HOME, 15, 'home')
11/08/2014 Carolina Bianchini 7 Calibration program improvements
11/08/2014 Carolina Bianchini 8 Calibration program improvements Now it’s possible to do the calibration because the program avoid problem: noise bigger then the peaks for a lot of position or not enough data
11/08/2014 Carolina Bianchini Conclusions 9 The time to run the program is fixed and actually the combination of sub routines is the fast one. I have introduced the control check for the bad measurement and impossibility to apply the gaussian fit; in order to use more calibration data. I have start to implement the compare calibration program, to check the sigma value between different calibrations, on the same wire scanner.
11/08/2014 Carolina Bianchini 10 Emiliano Thanks to...