The following event-related parameters are available for display/analysis through the screen numerical display, the event parameter list/save facility, the histogram facility, the 2- and 3-D scattergraph facilities, and the make analysis into data command.
Note that not all parameters are available in all facilities.
These parameters are intrinsic to the events themselves, and do not require measurements from any data trace.
Event ID: The identification number ( 1 - ...) of a particular event.
Time, On time: The onset time of the event, i.e. the acquisition time of the first datum within the event, where the acquisition time of the first datum in the file is 0.
Time (elap), On time (elap): The onset time of the event relative to elapsed experimental time.
For episodic files, this will be the time elapsed from the start of the episode plus the start time of the episode itself.
For non-episodic files, this will be the same as the simple On time parameter above.
On Duration, Duration: The duration of the event, i.e. the number of samples enclosed within the event multiplied by the sample interval.
Num samples: The number of data samples enclosed within the event.
Gap duration: Duration of the gap between two events, calculated as the time from the end of an event and the start of the following event. This is the same as the number of samples between the two events multiplied by the sample interval. Note that for identification purposes a gap is associated with the event that precedes it, so the start time of a gap is the start time plus duration of the associated preceding event. The gap duration of the last event is undefined.
Frequency (instantaneous): The instantaneous frequency (in Hz) of an event, calculated as the reciprocal of the time interval between the start of that event and the start of the following event. The frequency of the last event is undefined.
Frequency in time bin: The recording, or a section of the recording, is divided into a sequence of contiguous time bins. The number of events whose start times fall within each bin is divided by the bin width (in seconds) to calculate the average frequency of events within the bin. If chosen as the parameter for a Histogram display, the X axis represents contiguous time within the selected analysis region.
Count in time bin: The recording, or a section of the recording, is divided into a sequence of contiguous time bins. The number of events whose start times fall within each bin is calculated. If chosen as the parameter for a Histogram display, the X axis represents contiguous time within the selected analysis region.
Interval: The time interval between the start of an event and the start of the following event. For rhythmic data, this is the period of the rhythm. The interval of the last event is undefined.
Duty cycle: The fraction (0 - 1) that an event occupies in its current cycle. This is calculated as the event duration, divided by the time from the start of the event to the start of the next event (the cycle period or interval). The duty cycle of the last event is undefined.
Distribution count: This generates an event count distribution display, and is only available from the Event analysis: Histogram command. The selected region of data is divided into successive time bins of duration specified by the Distrib bin (ms) parameter visible in the Histogram dialog. The Histogram X axis is the count of events within a distribution bin length of data, and the Y axis shows the number of distribution bins containing that count of events. So the left hand bin shows how many distribution bin-lengths of data have 0 events within them, the 1-from-left shows how many distribution bin lengths have 1 event within them, etc.
Auto-correlogram: This applies to Histogram displays. It shows the number of events within each bin on the condition that there is an event in the same channel at time zero. It differs from the interval parameter in that it considers intervals between all pairs of events whose inter-event interval occurs within the time set by the X axis scales, irrespective of whether other events have intervened.
Tagged: Shows T (true) or F (false) to indicate whether the event has a tag.
Label: Shows the event label. Blank if no label.
Colour (hex value): Shows the event colour as a hexadecimal RGB value (order 0xbbggrr).
Surprise: The surprise value of events detected using the Poisson or Rank surprise method for burst detection. Only relevant for event channels populated using this method.
Phase (1 in 2): The phase of the start of each event in the target event (1) channel relative to the starts of the pair of surrounding events in the reference (2) channel. Thus phase represents the fraction of the time between two successive reference events at which an intervening target event occurs. Phase is greater than or equal to 0, and less than 1. Target events preceding the first or following the last reference event are ignored.
Phase (1 near 2): Phase of the start of each event in the target event (1) channel
relative to the closest start time of the pair of surrounding events in the base (2) channel. Target events preceding the first or following the last reference event are ignored.
Phase varies between +0.5 and -0.5. This has the same information content as above (Phase 1 in 2), but it prevents the wrap-around problem that occurs when target and base events are nearly synchronous.
PSTH (2->1): This generates a peri-stimulus time histogram, and is only available from the Event analysis: Histogram command. Events in channel 2 mark the stimulus and events in channel 1 mark the response to the stimulus. The response latency is calculated relative to the stimulus, and pre-stimulus time can be included as control data. This is similar to cross-correlogram (2->1), except that stimulus events must be further apart than the largest specified histogram range, so that each response corresponds to a unique stimulus. The facility also allows you to collate response counts on a per-stimulus basis, so that the mean and standard deviation of counts per stimulus can be displayed.
Latency (1 after 2): The latency from the start of each event in the reference (2) channel to following events in the target event (1) channel, up to the next reference event. Latency is greater than or equal to 0, and less than the interval between the reference event and the immediately following reference event (i.e. latency resets with each reference event).
Latency (1 from near 2): The latency from the start of each event in the target event (1) channel to the nearest event, either preceding or following, in the reference (2) channel. Latency can thus be either positive or negative.
There will always be the same number of latencies as there are events in the target channel.
Latency (1 from 2 paired): The latency from the start of each event in the target event (1) channel to the start of the event with the same ID number in the reference (2) channel. Latency can be either positive or negative.
Unpaired events are ignored.
Count (2 in 1): The count of events in the channel 2 that are entirely within the duration of events in the channel 1. A value is thus available for each event in channel 1. When displayed in a histogram, the X axis (bins) shows the number of enclosed channel 2 events, the Y axis shows the number of events in channel 1 which have that number of enclosed channel 2 events.
Stats (2 in 1): Show statistics of events in channel 2 which are entirely within the duration of events in channel 1. Statistics are thus available for each event in channel 1.
Count (2 after 1): The counts of events in channel 2 which follow after each event in channel 1. Counting is terminated when the off-time of a channel 2 event is later than the on-time of the next channel 1 event. When displayed in a histogram, the X axis (bins) shows the number of channel 2 events following a channel 1 event, the Y axis shows the number of events in channel 1 which have that number of following channel 2 events.
Cross-correlogram (2 -> 1): This applies to Histogram displays. It investigates the "influence" that events in a reference channel (channel 2) have on the probability of occurrence of events in a target channel (channel 1). Specifically, it shows the number of target events in channel 1 within each bin on the condition that there is an event in reference channel 2 at time zero. It thus shows the probability of occurrence of target events at various times relative to reference events. Effectively, peaks/troughs at positive latencies indicate the possibility of a connection from the reference to the target data source. It differs from the latency parameter in that it considers all events in the target event channel whose latency from the reference channel occurs within a time set by the X axis scales, irrespective of whether other reference channel events have intervened.
These parameters involve measuring values from the data in the parent trace of the event channel that are fully enclosed within each event timespan.
P2P amplitude: The difference between the maximum and minimum values (max - min). The data may be on a continuous slope, they do not have to have actual peaks or troughs.
P2P duration: Absolute value of the time intervals between the maximum and minimum data values.
MaxT - MinT: The time difference between the maximum and minimum values. May be positive (rising slope; minimum before maximum) or negative (falling slope; maximum before minimum). A signed version of P2P duration.
Max: The maximum value.
Min: The minimum value.
Average: The arithmetic mean of the values.
S.D.: The standard deviation of values.
Normal statistics (Avg, S.D): The average and S.D. as above.
Median: The median data value.
MAD: The median of the absolute deviations from the median.
Robust S.D.: The robust standard deviation is taken as the MAD multiplied by 1.4826. If the data follow a normal distribution then this yields the same value as the standard deviation.
Robust statistics (Median, robust S.D.): The median, MAD and robust standard deviation as above.
Area: The area is the integrated value, i.e. sum x sample interval.
Area re Base: The area (integrated values as above) relative to a baseline derived by linear interpolation between the first and last data values within the event. The first and last values thus do not contribute to the area.
Slope: The linear regression slope of data within the event, expressed as change in amplitude per second.
RMS: The root-mean-square of data within the event. I.e. square all the data values (thus making them >= 0), find the average (arithmetic mean) of these values, and report the square root of that average.
Energy (relative): The sum of the squared data values throughout the event, normalized to time by multiplying by the sample interval.
Auto Rhythm idx, Auto Rhythm period,
Autocorr rhythm info (rhythmicity idx, period): Perform a statistical autocorrelation (normalized autocovariance) analysis of the waveforms within each event, and determine the index of rhythmicity (the first trough to second peak height) and the period (lag to second peak).
PC 1-3 weight: Coefficient of the first, second or third principal component. If principal components have not previously been calculated then you will be asked whether you wish to calculate them.
Val at time A, B, C: The values of data at 3 different times within events. The times are specified in auxiliary program components.
DWT A, B, C: The value of the discrete wavelet transform coefficient of data within the event at times A, B and C. The times and wavelet characteristics are specified in auxiliary program components.
Template Error: The error between the waveform of the parent trace within this event and the template used in the Template recognition method of detecting events. Only relevant for event channels populated using this method.