Analysis Tasks
Selecting the File|Analyze... menu command raises a modal " task" dialog. Through this "wizard"-like dialog you can create, edit, and run analysis tasks that operate over some or all of the files in the current data directory. Currently, the only type of directory-wide analysis task supported is a trial data collation task. In this task, you specify the names of the trials you wish to process, the contiguous interval of the trial recording to be examined -- the analysis "epoch" --, and one or more data channels ("signals") for which data is to be extracted. There's a good deal of flexibility in specifying the analysis epoch, including the use of marker pulses. For each signal, you can choose to: (1) subtract a baseline level calculated over a specified window preceding the analysis epoch; (2) save either the entire collection of extracted traces or only the average computed over those traces; and (3) ignore any JMWork-marked saccades (eye-velocity channels only). For digital event and sorted-spike channel data, the "average" is actually a "sliding average" of event frequency (in Hz) over a specified window. For eye-velocity channels, sampled data within a marked saccade are treated as NaN if the saccade is marked with a mark1-mark2 pair, or are linearly interpolated between the saccade endpoints if the saccade is marked with a saccade-cut action.
Once you have defined a collation task, you can run it and review the collected results. The results panel includes a tab page for each distinct signal defined in the task. On the left-hand side is a file list and, below it, a text area that contains some information about the signal data collected from the data file highlighted in the list. On the right-hand side is a trace canvas. All of the collected signal traces are rendered here in gray; the trace extracted from the highlighted data file is in blue. If the signal average is to be saved rather than the collection, the average trace will be drawn in green, superimposed over the individual traces. Several keypress commands are available when this results panel is in front:
N, down arrow: Select the next data file in the results file list, wrapping back to the beginning of the list if necessary.
P, up arrow: Select the previous data file in the results file list, wrapping down to the end of the list if necessary.
X: Toggle the "exclude" flag for the selected data file. Excluded data files are marked in the file list by a red "X" icon. All data extracted from an excluded data file will be omitted if the results are saved and will not contribute toward the signal average (if applicable). Excluded data traces are drawn in red rather than gray; if the current trace is excluded, it is drawn in magenta instead of blue.
R: Reset the "exclude" flags for all data files in the results file list.
Once you have reviewed and are satisfied with the collation task results, press the Save Results button to save them to a DataNav-compatible source file. If you select an existing DataNav file, the task results will be appended to it. If you select a non-existent file, the results are saved in the binary format, which is the most efficient among the alternative DataNav file formats. The data can then be easily read into DataNav itself or imported into Matlab for further analysis using the M-function getdatanavsrc(), which is available from the DataNav online user's guide.
A user can define any number of analysis tasks. All task definitions are saved in the user's private JMWork directory, ready to be accessed the next time the user runs the program.