Advanced | SISEUIL

OVERSHOOT STATISTICS – [SISEUIL]
Compute overshoot statistics of a signal using a threshold condition.
CATEGORY
SI5

DESCRIPTION
This function is used to compute overshoot statistics using a threshold condition. It also generates a logical signal with these overshoot times on the signal studied.
RESULT VARIABLES AND PARAMETERS

Moyd	Sample mean computed on the first Time interval (Average Time Length) in milliseconds of the analyzed signal from the start of the signal. Afterwards, the signal overshoots are analyzed from the Starting time.
ttot	Sum of the overshoot times on the threshold value for the signal analyzed. This result is specified in milliseconds [ms].
tmax	Longest overshoot time specified in milliseconds [ms].
ncnt	Number of times when the threshold overshoot duration is longer than the specified Delay ON.
logi	The logical signal has a value of 1 from the start of an overshoot plus the Delay ON value. It remains at 1 until the end of the overshoot plus the specified Delay OFF value. For all other cases, the value of the logical signal is 0. The overshoot times which do not exceed the Delay ON value are not shown in the logical signal.
Threshold (+/-)	Difference with the mean of the starting samples to set the threshold value to reach. The threshold value used for overshoot calculations is the threshold difference added to the mean result (the mean of the time interval samples from the start of the study). This difference is expressed in the units of the signals. Example: Let, the mean moyd = 40.0. If the threshold difference is = 15.0, then the threshold value used is 55.0. If the threshold difference is = -15.0, then the threshold value used is 25.0. The threshold direction is set by the sign of the threshold. If the difference is negative, then the threshold value is smaller than moyd. Hence, overshoots going towards the bottom part of the graph are looked for, since the start of the study is not considered as an overshoot. This case is shown in first example below. If the difference is positive, then the threshold value is greater than moyd. Hence, overshoots going towards the top part of the graph are looked for, since the start of the study is not considered as an overshoot. This case is shown in the second example. To be considered as an overshoot, the signal must reach the threshold and return under this threshold within the set limits of the analysis. Thus, if a signal is above a threshold throughout the analysis, then the ncnt, ttot and tmax results will be null.
Average Time Length	Time interval in milliseconds [ms] from time 0 on which the initial mean will be computed and serving to set the threshold. A null value is not valid.
Delay ON	Minimal time during which the overshoot must last in order to be considered as a count. This time is specified in milliseconds [ms].
Delay OFF	Minimal time during which the logical signal generated must stay at 1 following the end of an overshoot. This time is specified in milliseconds [ms].
Begin_time	Time at which the analysis of a signal must start. This time is expressed in milliseconds [ms]. This value must be greater than 0 and lower than the duration of the test. The default value is 0.
End_time	Time at which the analysis of a signal must end. This time is expressed in milliseconds [ms]. The value of this time must be larger than the specified begin_time and smaller than the duration of the test. Use the “HUGE” value to specify the end of the test. The default value is HUGE. The begin_time and end_time specified for the calculation of the harmonics must be the same as those used in the sifreq sequences to compute the fundamental frequency.
Factor	Multiplying factor for the results generated by the function. The default value of the multiplying factor is 1.0 and has no effect on the results. The factor applies to the moyd, tmax and ttot results.

SYNTAX

[moyd, ttot, tmax, ncnt, logi] = siseuil( input, 0.5, 1.0, 0.5, 0.25, 0, HUGE, 1)

CHARACTERISTICS
Data type support
Double Floating point
EXAMPLE
The first figure shows a case where the threshold difference is negative.

The second figure shows a case where the threshold difference is positive.