Cross-calibration

The cross-calibration tool on this page was used in the construction of the SEPEM reference proton dataset. It can be applied to channels of data tables created by users.

Table and channel selection

First select the data table to be used as a baseline for the calibration (table 1), and the data table which should be re-calibrated (table 2). When both tables have been selected, the page reloads with a selection of the data channels.

When the channels for tables 1 and 2 have been selected, press the Run button to start the calculation of the calibration function.

The calibration procedure starts by building a list of epochs which coincide for both data tables. If there are no matching epochs, the process is aborted and a warning message is displayed. If at least two matching epochs are found, a linear regression is calculated of the form

Y = a + b X ,
where X represents channel 1 and Y represents channel 2. The actual cross-calibration is achieved by applying the inverse relation to channel 2.

Results

The regression function is printed with the calculated regression coeffients and the number of matching data points. Four plots are shown as well:
  1. the original data on a linear scale, with the regression line and the bisector superimposed;
  2. the original data on a logarithmic scale, with the regression line and the bisector superimposed;
  3. the cross-calibrated data on a linear scale, with the original regression line and the bisector superimposed;
  4. the cross-calibrated data on a logarithmic scale, with the original regression line and the bisector superimposed;
If the current user is the creator of table 2 (the table to be cross-calibrated), the calibration relation can be applied to channel 2 by pressing the Apply Calibration button (which is otherwise invisible). This action will update the data for channel 2 in the database.
Caution: This action is irreversible!




Last modified on: 3 October 2011.