How to use calib DB module

The following example shows how to use the calib_DB class, using an acceptance tests measurement (60 seconds measurement with the DOM oriented in each of the 4 cardinal points).

This calibration module will use the module ID to retrieve the proper compass calibration in the km3 central DB.

import km3compass as kc
import matplotlib.pyplot as plt

Loading some data

Initialising is a simple as this:

filename = "../tests/DOM_0801.csk"
reader = kc.readerCSK(filename)
print(reader.module_IDs)

File loaded, 2508 rows
        1 module(s)
        - 817302522
Number of measurements after removing duplicates : 252
[817302522]

Applying calibration

The calib_DB is a quite easy object to handle. It expects as input a python class with a property .df, that contains a pandas.DataFrame with columns : ['hx','hy','hz','ax','ay','az','moduleID']. Also, this module expects only one module, so you should provide the the module ID you want to use.

calib = kc.calib_DB(reader, 817302522)

DOM mac address : 08:00:30:b7:0b:fa

Print calibration information

Data (both acceleration and magnetic field) are calibrated by first applying an offset (3D vector) to the data, followed by a rotation matrix (3x3 matrix). Both of this elements can be displayed using embeded functions.

calib.print_calibration()

----------------------------------------
Calibration for compass 845
Type: "AHRS-CALIBRATION-v3", source: "DB"
A norm = 1.0
A xyz offsets = [ 0.013932  -0.0056594 -0.0046136]
A rotation matrix =
[[ 1.0175    -0.14534   -0.0062442]
 [-0.14534    0.99582    0.0069356]
 [-0.0062442  0.0069356  1.008    ]]
H norm = 1.0
H xyz offsets = [ 0.02418  -0.076915  0.023885]
H rotation matrix =
[[ 0.97688    0.0088185 -0.0040336]
 [ 0.0088185  1.0095    -0.007019 ]
 [-0.0040336 -0.007019   1.0141   ]]
----------------------------------------

Comparing data before and after calibration

To do so, we will use the kc.plot_raw_results function, that works for both raw and calibrated data :

kc.plot_raw_results(reader.df, title="Data before calibration")

kc.plot_raw_results(calib.df, title="Data after applying DB calibration")

• 
• 

<Figure size 1600x700 with 6 Axes>

Comparing the results before and after calibration, we can already see that the magnetic field seems now to be centered around a point close to 0, which is what we expect. In the case of the acceptance tests, we know that the rotation should be only the XY plan but we can see that some of the movement seems to be along the Z direction. Looking at the right most plot, we can see that the acceleration is not fully aligned with the Z axis (i.e. angle is not 90°). To correct this, we can manually correct this using the accelerometer information.

Aligning Z direction with weight

**Strictly speaking, this part is not a calibration, but more a correction. Don’t use it blindly, only if it makes sense !**Schematically, what we want is to rotate all the measurement to have the weight (i.e. full acceleration) aligned with the Z axis. To do so, a function is provided in km3compass : kc.align_z2weight

df = kc.align_z2weight(calib.df)

kc.plot_raw_results(
    df, title="Data after applying calibration and aligning weight with Z"
)


plt.show()