reference

`nerve_response_analysis.analysisRun.analyzeSession`

Analyze accepted RHS recordings and combine their results.

Syntax

analysis = nerve_response_analysis.analysisRun.analyzeSession(session)
analysis = nerve_response_analysis.analysisRun.analyzeSession( ...
session, protocol)
analysis = nerve_response_analysis.analysisRun.analyzeSession( ...
session, protocol, opts)

Description

Selects accepted rows from an RHS Preview filter record, analyzes each file with analyzeRecording, and vertically concatenates event, train, metric, and issue tables. A failure in one recording becomes an issue row and does not stop later recordings.

Inputs

session: Scalar structure whose recordings field is a table or structure array. Required recording columns are recordingId and filePath. label, comment, qcFlag, and keep are optional selection/provenance fields.
protocol: Optional channel-role and pair protocol passed unchanged to each recording. See analyzeRecording. Default: struct().
opts: Optional scalar structure. Recording-level fields are forwarded to analyzeRecording after recordingId is replaced by the current row id.

Recording Selection

If keep exists, rows with logical keep=true are selected. Otherwise label selects normalized "good" rows; accepted synonyms are good, keep, kept, true, 1, yes, and y, while bad/reject synonyms normalize to "bad". If label is absent but qcFlag exists, rows equal to "accepted" are selected. With no selection column, every row is treated as good. This precedence means keep overrides label and qcFlag when more than one is present.

Options

maxRecordings: Positive finite limit applied after recording selection. Zero, nonpositive, nonfinite, or omitted input means no limit. Default: Inf.
maxDurationSec: Forwarded to analyzeRecording for every selected file.
eventDetection: Forwarded to analyzeRecording for every selected file.

Outputs

analysis: Scalar session-analysis structure.

Analysis Fields

type, version - "nerveResponseSessionAnalysis" and schema version 1.

recordingCount: Total rows before acceptance filtering.
analyzedCount: Number of selected rows after maxRecordings. This counts attempted recordings, including files that later produce issues.
protocolSnapshot: Protocol supplied by the caller. events, trains, metrics - Combined nonempty tables returned by recording analyses.
issues: Combined recording issues plus caught exceptions, with recordingId, severity, and message columns.

Failure Behavior

A recording-level analysis exception is converted to one issue row and later recordings continue. Missing or unsupported session.recordings input becomes an empty session; a nonempty table without recordingId or filePath, incompatible selection columns, or unappendable result tables may still raise the originating table/field error before or between recordings.

Example

recordings = table(strings(0,1), strings(0,1), ...
'VariableNames', {'recordingId', 'filePath'});
session = struct("recordings", recordings);
analysis = nerve_response_analysis.analysisRun.analyzeSession(session);
assert(analysis.recordingCount == 0 && analysis.analyzedCount == 0)

nerve_response_analysis.analysisRun.analyzeRecording — Analyze one RHS recording according to a protocol.
nerve_response_analysis.analysisRun.detectEventTrains — Detect pulse candidates and group them into trains.
nerve_response_analysis.analysisRun.measureCapMetrics — Measure event-locked CAP response features.

Source

This page is generated from the MATLAB help text in apps/neurophysiology/nerve_response_analysis/+nerve_response_analysis/+analysisRun/analyzeSession.m.

nerve_response_analysis.analysisRun.analyzeSession