Trial controls the running of trials and sets of trials.

some_trial add_stimulus_event name slot start end value...

Add a stimulus clamping event named name to the trial sequence of some_trial. The stimulus will be clamped to the input of the model with the name slot, which should be a conventional name (e.g., orthographic) that abstracts away from particular trials and models. The elements in value (which may be macro variable names preceded by $) will be interpreted via the representation associated with that input slot to be converted to the representation for the corresponding input layer. Add start as a starting trigger event for this clamping event; add end as an ending trigger event for this clamping event. Both start and end are with the syntax described at add_stimulus_start_condition.

The onset and offset of the clamping are given by start and end which are of the format event+delay where delay is a number (or $macro variable indicating the amount of time to elapse after event for this to occur (+0 may be omitted); event may be BEGIN, END, obsname, or ename/start or ename/end, where obsname is the name of an observer, or ename is the name of another stimulus event added to the trial.

some_trial add_stimulus_event_only name

Add a stimulus clamping event named name to the trial sequence of some_trial, without setting any of the properties of this event.

some_trial set_description blah...

Associate a description blah with some_trial.

some_trial set_stimulus_event_slot name slot

Identify the slot name that determines which layer and representation are used for clamping the event name of some_trial to be slot. The stimulus will be clamped to the input of the model with the name slot, which should be a conventional name (e.g., orthographic) that abstracts away from particular trials and models.

some_trial set_stimulus_event_value name value...

Set the list of values that are used for clamping event name of some_trial to be value. The elements in value will be interpreted via the representation associated with the prevailing input slot to be converted to the representation for the corresponding input layer.

some_trial remove_stimulus_event name

Remove a stimulus clamping event named name from the trial sequence of some_trial.

some_trial add_stimulus_start_condition eventname starttime

Indicate that starttime is one of the possible trigger times for the clamping event eventname of some_trial to onset; events always start when the first such time is reached. The format of starttime is event+delay where delay is a number (or $macro variable) indicating the amount of time to elapse after event for this to occur (+0 may be omitted). event may be literally BEGIN or END; the name of an observer; or ename/start or ename/end, where ename is the name of another stimulus event added to the trial.

some_trial add_stimulus_end_condition eventname endtime

Indicate that endtime is one of the possible trigger times for the event eventname of some_trial to offset; events always start when the first such time is reached. The format of endtime is as for starttime above.

some_trial remove_stimulus_start_condition eventname starttime

Remove the onset tigger starttime for the event eventname of some_trial.

some_trial remove_stimulus_end_condition eventname endtime

Remove the offset condition endtime for the event eventname of some_trial.

some_trial set_end_condition terminationtime

Indicate that the trial some_trial ends when the termination time/event terminationtime is reached. The format of terminationtime is as for startime of add_stimulus_start_condition.

some_trial remove_end_condition terminationtime

Delete terminationtime from the list of times at which some_trial will end.

some_trial details

Emit an extensive description of events and end conditions.

some_trial default_observer

Return the dflog_observer that is managed by some_trial and whose settings (by default) record the content of any macro variables and the time that each observer mentioned in the trial description (first) becomes true.

some_trial set_cap timeout

Establish a special timeout termination condition for some_trial. timeout is a number (without leading + symbol). Layers will not be advanced beyond this time, and once all layers reach timeout, NAN (interpreted as timeout) will be recorded as the time of all observers that have been false throughout (unless overridden by set_cap_validity).

some_trial event_pattern name

Return the pattern object that is used for clamping the event name during the trial some_trial.

some_trial response name

Return a string pattern that represents the state of the observer name when it (first) becomes true during some_trial. Only valid for observers named as a time/event for clamping start/end or trial end.

some_trial response_time name

Return a strength pattern that represents the time at which the observer name when it (first) becomes true during some_trial. Only valid for observers named as a time/event for clamping start/end or trial end.

some_trial run_set df_stimset

For each row in the dataframe/stimulus_set df_stimset, run some_trial with local macro variables with the names of the columns of the df_stimset set to the corresponding values on the row. The default_observer is set up and enabled with appropriate trigger before the run, and this process is reversed after the run.

some_trial run_set_multiple spec num df_stimset

As run_set except that each row is used multiple times, the number of which is set by num, and the ordering of which is set by spec. A spec of ABAB indicates that the whole set should be run through before the next of the num repetitions, whereas AABB indicates that that each row should be run num times before proceeding to the next row.

some_trial set_cap_validity 0

Indicate that when the cap timeout is reached, the time, not NAN (timeout), should be recorded as response_time for observers that have no become true.

some_trial set_cap_validity 1

Revert to the normal behaviour of the cap, where NAN (timeout) is recorded.