Back to the main page.
Bug 1141 - extend the ft_analysisprotocol function with reports on the callinfo
| Status | CLOSED FIXED |
| Reported | 2011-11-11 11:07:00 +0100 |
| Modified | 2012-03-14 10:00:41 +0100 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | All |
| Operating System: | All |
| Importance: | P3 enhancement |
| Assigned to: | Eelke Spaak |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Robert Oostenveld - 2011-11-11 11:07:58 +0100
Besides fieldtrip version, we are currently also recording the callinfo with matlab version, user, time and mem use, etc. It would be nice to be able to also report those in ft_analysisprotocol, e.g. like cfg.showmatlabversion = yes, no cfg.showcomupter = yes, no cfg.showuser = yes, no cfg.showfunctionname = yes, no this would extend the boxes vertically, and print the additional information that was tracked in the cfg.previous.previous... and at the end include a summary like fprintf('the recorded analysis pipeline took in total %f hours to run', ...); fprintf('the recorded maximum memory requirement of the analysis pipeline was %d GB', ...);
Boris Reuderink - 2011-11-17 10:46:46 +0100
Changed the status of bugs without a specific owner to UNCONFIRMED. I'll try to replicate these bugs (potentially involving the submitter), and change confirmed bugs to NEW. Boris
Boris Reuderink - 2011-11-17 13:31:55 +0100
Is this related to bug #1142?
Robert Oostenveld - 2011-11-17 14:15:01 +0100
(In reply to comment #2) yes, related but neither is really dependent on the other. The history has to be fully kept (1142) and the history should be nicely visualized (1141)
Eelke Spaak - 2011-12-13 14:55:25 +0100
I have committed the fix (revision 5028), but it might be wise if Robert or someone else tests the function with all its possible uses. I did not really know how to test the branching functionality, i.e. when does an analysis pipeline branch? (Depth is fine and tested by myself.) The entire algorithm for plotting the gui needed to be rewritten to accomodate flexibly sized boxes, so testing is probably a good idea.
Robert Oostenveld - 2011-12-13 16:08:32 +0100
(In reply to comment #4) branching (when going upward, or merging when going downward) happens in functions that combine multiple inputs. E.g. grandaverage, or statistics. In that case cfg.previous = {varargin{1}.cfg, varargin{2}.cfg, ...} A new application for branching/merging is now also with sideways entry of previous cfgs in volcfg = ... vol = ft_prepare_headmodel(volcfg) sourcecfg = ... sourcecfg.vol = vol; % note that this also contains the volcfg source = ft_sourceanalysis(cfg, data) where the output source.cfg = data.cfg, but where source.cfg.vol.cfg is a "sideways" branch into the headmodel construction tree (which can contain a sequence of volumerealign, volumereslice, and volumesegment prior to ft_prepare_headmodel). The visualisation of the sideways branching would probably best be done with something like fn = fieldnames(cfg) for i=1:length(fn) if isa(cfg.(fn{i}), 'struct') && isfield(cfg.(fn{i}), 'cfg') % sideways branch detected, do something with it end Could you still add this to the ft_analysisprotocol visualisation?
Eelke Spaak - 2012-01-13 11:33:40 +0100
Changing this to resolved again, to keep bugs tidy (one bug == one issue).