Magic Numbers

There are several places in the cait code, where so-called “magic” numbers appear, e.g. a hard-coded integer as index to an array, with no or only short explanation why this number is taken here. Most of the time, this is due to the numbering of labels, main parameters or fit parameters. These numbers are written as attributes in the HDF5 files, however we also provide lists here.

Warning

Note that the numbers for main parameters and additional main parameters only apply for those that were calculated with the (now deprecated) functions dh.calc_mp and dh.calc_additional_mp. You should start using dh.cmp, which saves datasets with descriptive names such that we will not need these “magic” numbers anymore in the future.

Main Parameters

array index	name	meaning
0	pulse_height	The Maximum Value of the Event, after application of a 50 sample moving average, in Volt.
1	t_zero	This value is calculated backwards in time: Starting at the position of the sample with the maximum value, the last sample that subceeds 20% of the pulse height.
2	t_rise	This value is calculated backwards in time: Starting at the position of the sample with the maximum value, the last sample that subceeds 80% of the pulse height.
3	t_max	The index of the maximal sample.
4	t_decaystart	The first sample after t_max, that falls below 90% of the maximal height.
5	t_half	The first sample after t_max, that falls below 73% of the maximal height.
6	t_end	The first sample after t_max, that falls below 36% of the maximal height.
7	offset	The average of the first 500 samples of the event.
8	linear_drift	The difference between the average of the first and last 500 samples of the event, divided by the record length.
9	quadratic_drift	Usually this is set to zero! The quadratic component of the baseline.

Additional Main Parameters

array index	name	meaning
0	array_max	The maximum of the array.
1	array_min	The minimum of the array.
2	var_first_eight	The variance of the first eight of the record window. This is typically the variance of the baseline noise.
3	mean_first_eight	The mean value of the first eight of the record window. The first eight of the record window typically shows only the noise baseline, i.e. this is the mean value of the noise.
4	var_last_eight	The variance of the last eight of the record window. If this value differs strongly from the baseline variance, this indicates a strongly saturated pulse, a Pile-Up Event or an early trigger.
5	mean_last_eight	The mean value of the last eight of the record window. A difference of this value from the mean of the first eight indicates a strong baseline tilt.
6	var	The variance of the whole array.
7	mean	The mean value of the whole array.
8	skewness	The skewness of the whole array.
9	max_derivative	The maximal value of the derivative of the array.
10	ind_max_derivative	The index of the maximal value of the derivative of the array.
11	min_derivative	The minimal value of the derivative of the array.
12	ind_min_derivative	The index of the minimal value of the derivative of the array.
13	max_filtered	The maximum of the array, after applying the optimum filter.
14	ind_max_filtered	The index of the maximum of the array, after applying the optimum filter.
15	skewness_filtered_peak	The skewness of the array around its peak, after applying the optimum filter. Typically this value is higher, when the array deviates from the standard event.

Parametric Fit Parameters

The fit parameters of the parametric pulse shape fit.

array index	name	meaning
0	\(t_0\)	onset of the pulse
1	\(A_n\)	non-thermal amplitude
2	\(A_t\)	thermal amplitude
3	\(\tau_n\)	non-thermal time constant
4	\(\tau_{in}\)	intrinsic TES time constant
5	\(\tau_t\)	thermal time constant

Standard Event Fit Parameters

The fit parameters of the standard event fit:

array index	meaning
0	pulse height
1	onset
2	constant coefficient
3	linear coefficient
4	quadratic coefficient
5	cubic coefficient

The array fit has the same parameters!

Labels

label	name	meaning
0	unlabeled
1	Event Pulse	A usual pulse, originating in a particle recoil within the absorber crystal.
2	Test/Control Pulse	A testpulse, induced by a temperature pulse from the detector heater. These pulses are intentionally induced every few seconds, to measure the variations in the detector response. Their pulse form is different than that of event pulses, with typically a longer rise time.
3	Noise	An empty noise baseline, or a noise trigger.
4	Squid Jump	A jump in the offset level of the baseline, coming from a reset of the SQUID or the loss of a flux quant.
5	Spike	A thin positive or negative spike from the detector electronics.
6	Early or late Trigger	An absorber pulse with maximum time significantly different than zero (>/< 20 ms).
7	Pile Up	Two or more events in one record window.
8	Carrier Event	A thin, low energetic pulse, originating in a particle recoil within the carrier crystal of the detector.
9	Strongly Saturated Event Pulse	Saturated Event Pulse. Saturation happens, when the TES is driven out of its linear operation range by a high energetic particle recoil within the absorber crystal.
10	Strongly Saturated Test/Control Pulse	Same as Label 9, but for a Test Pulse.
11	Decaying Baseline	Trigger of noise or a sub-threshold event, that gets elevated above threshold by a decaying baseline.
12	Temperature Rise	A sudden and continuous rise of the detector temperature causes a strong positive slope in the noise baseline, that eventually surpasses the threshold.
13	Stick Event	A pulse with long rise- and decay time, caused by a particle recoil within the sticks that hold the crystal. The sticks themselves are made from e.g. CaWO4.
14	Square Waves	Small, discrete jumps of the baseline level. A upward jump, followed right away by a downward jump, is also called a jump event and might look like a pulse on first sight.
15	Human Disturbance	An oscillating signal, caused by vibration.
16	Large Sawtooth	Multiple squid resets, cause by fast rising temperature - e.g. due to warm up of the cryostat or very large heat deposition to a sensible TES.
17	COSINUS Tail	An event with especially long tail, typical for the several COSINUS modules.
18	Light only Event	A direct hit in the light detector, without corresponding phonon signal.
19	Ring & Light Event	A recoil in the ring or the detector, this component is e.g. included in Gode/beaker modules.
20	Sharp Light Event	A very fast rising light event, typically caused by a direct hit in the light detector, with a hit in the detector following after.
99	unknown/other	In some cases, we have not enough insight to explain an event. If this concerns only a single or very few event, these are together put in an unknown class, which is then excluded from the analysis.