Back to the main page.
Bug 216 - to do: develop ft_qualitycheck
| Status | CLOSED FIXED |
| Reported | 2010-11-12 13:37:00 +0100 |
| Modified | 2011-03-31 14:13:58 +0200 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Windows |
| Importance: | P1 normal |
| Assigned to: | Arjen Stolk |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Arjen Stolk - 2010-11-12 13:37:17 +0100
FT_QUALITYCHECK should read data and visualize the quantification of its quality There will be 3 steps for a user to check his/her data. 1) create an ouput.mat file with the quantified data (which ft_qualitycheck will do) 2) visualize an overview of the results (which ft_qualitycheck will do) 3) a more detailed inspection (user, and some examples on the WIKI) to do list of adjustments/implementations: - powspctrm: 0-400 hz, loglog plot - remove pie's and possibly add day of the week - min/maxima: change to mean and range, fixed y-axis - timelimits: max 1 hour, 2nd plot when exceeding - linenoise: take ratio, semilogy + grid plot - lowfreq noise: semilogy + grid plot - topo: colormap dutch, based on artifacts where max-min'ish > something, plot the sum in stead of the mean per 10 secs, interpmethod nearest - topoREF: imsagesc, matrix-like in stead of 'organised' - update document help with the user steps - triggers found per chan - timeaxis should be 5,10,15,etc. secs - also show which channels are missing from the analysis - .mat file should behave according to a timelock analysis structure
Arjen Stolk - 2010-11-28 16:41:00 +0100
done: - create headposition and headmotion variables - visualize headmotion - powspctrm: 0-400 hz, loglog plot - output.mat - .mat file should behave according to a timelock/freq analysis structure - update document help with the user steps - timeaxis should be 5,10,15,etc. secs - linenoise: take ratio, semilogy + grid plot - lowfreq noise: semilogy + grid plot - min/maxima: change to mean and range, fixed y-axis - remove pies and possibly add day of the week - topo: colormap dutch, sum of artifacts instead of mean per trial, interpmethod nearest yet still to implement: - timelimits: max 1 hour, 2nd plot when exceeding - topoREF: imsagesc, matrix-like in stead of 'organised': give sensornames with > 0 artifacts, remove scale etc. - triggers/events found per chan - show which channels are missing from the recording - cfg.input- and outputfile compatibility
Arjen Stolk - 2010-12-17 12:53:05 +0100
done: - 10 second data is redefined in 1 second pieces. This speeds up the freqanalysis and reduces required storage (size .mat file). - create headposition and headmotion variables - visualize headmotion - powspctrm: 0-400 hz, loglog plot - output.mat - .mat file should behave according to a timelock/freq analysis structure - update document help with the user steps - timeaxis should be 5,10,15,etc. secs - linenoise: take ratio, semilogy + grid plot - lowfreq noise: semilogy + grid plot - min/maxima: change to mean and range, fixed y-axis - remove pies and possibly add day of the week - topo: colormap dutch, sum of artifacts instead of mean per trial, interpmethod nearest yet still to implement: - chronjob ft_qualitycheck at Odin - remove quantification sliders; space may be used for buttons that 'zoom in' - topoplot colorbar should be white -> red - headmotion: barplot -> boxplot (second from below in doc page, ylim[0 10] mm, change coilnames to N,L,R - timelimits: max 1 hour, 2nd plot when exceeding - topoREF: imsagesc, matrix-like in stead of 'organised': give sensornames with > 0 artifacts, remove scale etc. - triggers/events found per chan - show which channels are missing from the recording - cfg.input- and outputfile compatibility
Arjen Stolk - 2011-01-12 18:06:52 +0100
done: - 10 second data is redefined in 1 second pieces. This speeds up the freqanalysis and reduces required storage (size .mat file). - create headposition and headmotion variables - visualize headmotion - powspctrm: 0-400 hz, loglog plot - output.mat - .mat file should behave according to a timelock/freq analysis structure - update document help with the user steps - timeaxis should be 5,10,15,etc. secs - linenoise: take ratio, semilogy + grid plot - lowfreq noise: semilogy + grid plot - min/maxima: change to mean and range, fixed y-axis - remove pies and possibly add day of the week - topo: colormap dutch, sum of artifacts instead of mean per trial, interpmethod nearest - headmotion: barplot -> boxplot (second from below in doc page, ylim[0 10] mm, change coilnames to N,L,R - topoplot colorbar should be white -> red ('hot') - topoREF: imsagesc, matrix-like in stead of 'organised': give sensornames with > 0 artifacts, remove scale etc. - show which channels are missing from the recording yet still to implement: - chronjob ft_qualitycheck at Odin - remove quantification sliders; space may be used for buttons that 'zoom in' - timelimits: max 1 hour, 2nd plot when exceeding - triggers/events found per chan - cfg.input- and outputfile compatibility
Arjen Stolk - 2011-01-13 18:17:25 +0100
done: - 10 second data is redefined in 1 second pieces. This speeds up the freqanalysis and reduces required storage (size .mat file). - create headposition and headmotion variables - visualize headmotion - powspctrm: 0-400 hz, loglog plot - output.mat - .mat file should behave according to a timelock/freq analysis structure - update document help with the user steps - timeaxis should be 5,10,15,etc. secs - linenoise: take ratio, semilogy + grid plot - lowfreq noise: semilogy + grid plot - min/maxima: change to mean and range, fixed y-axis - remove pies and possibly add day of the week - topo: colormap dutch, sum of artifacts instead of mean per trial, interpmethod nearest - headmotion: barplot -> boxplot (second from below in doc page, ylim[0 10] mm, change coilnames to N,L,R - topoplot colorbar should be white -> red ('hot') - topoREF: imsagesc, matrix-like in stead of 'organised': give sensornames with > 0 artifacts, remove scale etc. - show which channels are missing from the recording yet still to implement: - BTI-compatibility - patching 10% of maxima and minima around the mean, for example with this line of code: h = patch([time time(end:-1:1) time(1)], [avg_min avg_max(end:-1:1) avg_min(1)], 'HandleVisibility', 'Off'); uistack(h, 'bottom'); where avg_min = avg - minima /10 and avg_max = avg + maxima /10 - change linenoise ratio into linenoise absolute raw power. This speeds up the analysis process as freqanalysis can be called one time less. - change legend, titles etc (e.g. 'headmotion from start') - chronjob ft_qualitycheck at Odin - remove quantification sliders; space may be used for buttons that 'zoom in' - timelimits: max 1 hour, 2nd plot when exceeding - triggers/events found per chan - cfg.input- and outputfile compatibility
Arjen Stolk - 2011-03-02 21:04:20 +0100
done: - 10 second data is redefined in 1 second pieces. This speeds up the freqanalysis and reduces required storage (size .mat file). - create headposition and headmotion variables - visualize headmotion - powspctrm: 0-400 hz, loglog plot - output.mat - .mat file should behave according to a timelock/freq analysis structure - update document help with the user steps - timeaxis should be 5,10,15,etc. secs - linenoise: take ratio, semilogy + grid plot - lowfreq noise: semilogy + grid plot - min/maxima: change to mean and range, fixed y-axis - remove pies and possibly add day of the week - topo: colormap dutch, sum of artifacts instead of mean per trial, interpmethod nearest - headmotion: barplot -> boxplot (second from below in doc page, ylim[0 10] mm, change coilnames to N,L,R - topoplot colorbar should be white -> red ('hot') - topoREF: imsagesc, matrix-like in stead of 'organised': give sensornames with > 0 artifacts, remove scale etc. - show which channels are missing from the recording - BTI-compatibility - change linenoise ratio into linenoise absolute raw power. This speeds up the analysis process as freqanalysis can be called one time less. - change legend, titles etc (e.g. 'headmotion from start') - patching 10% of maxima and minima around the mean, for example with this line of code: h = patch([time time(end:-1:1) time(1)], [avg_min avg_max(end:-1:1) avg_min(1)], 'HandleVisibility', 'Off'); uistack(h, 'bottom'); where avg_min = avg - minima /10 and avg_max = avg + maxima /10 yet still to implement: - chronjob ft_qualitycheck at Odin - remove quantification sliders; space may be used for buttons that 'zoom in' - timelimits: max 1 hour, 2nd plot when exceeding - triggers/events found per chan - cfg.input- and outputfile compatibility
Arjen Stolk - 2011-03-17 12:14:49 +0100
done: - 10 second data is redefined in 1 second pieces. This speeds up the freqanalysis and reduces required storage (size .mat file). - create headposition and headmotion variables - visualize headmotion - powspctrm: 0-400 hz, loglog plot - output.mat - .mat file should behave according to a timelock/freq analysis structure - update document help with the user steps - timeaxis should be 5,10,15,etc. secs - linenoise: take ratio, semilogy + grid plot - lowfreq noise: semilogy + grid plot - min/maxima: change to mean and range, fixed y-axis - remove pies and possibly add day of the week - topo: colormap dutch, sum of artifacts instead of mean per trial, interpmethod nearest - headmotion: barplot -> boxplot (second from below in doc page, ylim[0 10] mm, change coilnames to N,L,R - topoplot colorbar should be white -> red ('hot') - topoREF: imsagesc, matrix-like in stead of 'organised': give sensornames with > 0 artifacts, remove scale etc. - show which channels are missing from the recording - BTI-compatibility - change linenoise ratio into linenoise absolute raw power. This speeds up the analysis process as freqanalysis can be called one time less. - change legend, titles etc (e.g. 'headmotion from start') - patching 10% of maxima and minima around the mean, for example with this line of code: h = patch([time time(end:-1:1) time(1)], [avg_min avg_max(end:-1:1) avg_min(1)], 'HandleVisibility', 'Off'); uistack(h, 'bottom'); where avg_min = avg - minima /10 and avg_max = avg + maxima /10 - chronjob ft_qualitycheck at Odin - output .mat file has info, timelock, freq, summary and headpos structure - timecourses of headcoil distances - cleaned up code yet still to implement: - chronjob correct 'exportname' - remove quantification sliders; space may be used for buttons that 'zoom in' - timelimits: max 1 hour, 2nd plot when exceeding - triggers/events found per chan - cfg.input- and outputfile compatibility
Arjen Stolk - 2011-03-17 13:46:51 +0100
done: - 10 second data is redefined in 1 second pieces. This speeds up the freqanalysis and reduces required storage (size .mat file). - create headposition and headmotion variables - visualize headmotion - powspctrm: 0-400 hz, loglog plot - output.mat - .mat file should behave according to a timelock/freq analysis structure - update document help with the user steps - timeaxis should be 5,10,15,etc. secs - linenoise: take ratio, semilogy + grid plot - lowfreq noise: semilogy + grid plot - min/maxima: change to mean and range, fixed y-axis - remove pies and possibly add day of the week - topo: colormap dutch, sum of artifacts instead of mean per trial, interpmethod nearest - headmotion: barplot -> boxplot (second from below in doc page, ylim[0 10] mm, change coilnames to N,L,R - topoplot colorbar should be white -> red ('hot') - topoREF: imsagesc, matrix-like in stead of 'organised': give sensornames with > 0 artifacts, remove scale etc. - show which channels are missing from the recording - BTI-compatibility - change linenoise ratio into linenoise absolute raw power. This speeds up the analysis process as freqanalysis can be called one time less. - change legend, titles etc (e.g. 'headmotion from start') - patching 10% of maxima and minima around the mean, for example with this line of code: h = patch([time time(end:-1:1) time(1)], [avg_min avg_max(end:-1:1) avg_min(1)], 'HandleVisibility', 'Off'); uistack(h, 'bottom'); where avg_min = avg - minima /10 and avg_max = avg + maxima /10 - chronjob ft_qualitycheck at Odin - output .mat file has info, timelock, freq, summary and headpos structure - timecourses of headcoil distances - cleaned up code - updated help - timelimits: max 1 hour, 2nd plot when exceeding not implemented: - remove quantification sliders; space may be used for buttons that 'zoom in' - cfg.input- and outputfile compatibility - triggers/events found per chan yet still to implement: - chronjob should match with ft_qualitycheck 'exportname'
Arjen Stolk - 2011-03-18 15:40:34 +0100
done: - 10 second data is redefined in 1 second pieces. This speeds up the freqanalysis and reduces required storage (size .mat file). - create headposition and headmotion variables - visualize headmotion - powspctrm: 0-400 hz, loglog plot - output.mat - .mat file should behave according to a timelock/freq analysis structure - update document help with the user steps - timeaxis should be 5,10,15,etc. secs - linenoise: take ratio, semilogy + grid plot - lowfreq noise: semilogy + grid plot - min/maxima: change to mean and range, fixed y-axis - remove pies and possibly add day of the week - topo: colormap dutch, sum of artifacts instead of mean per trial, interpmethod nearest - headmotion: barplot -> boxplot (second from below in doc page, ylim[0 10] mm, change coilnames to N,L,R - topoplot colorbar should be white -> red ('hot') - topoREF: imsagesc, matrix-like in stead of 'organised': give sensornames with > 0 artifacts, remove scale etc. - show which channels are missing from the recording - BTI-compatibility - change linenoise ratio into linenoise absolute raw power. This speeds up the analysis process as freqanalysis can be called one time less. - change legend, titles etc (e.g. 'headmotion from start') - patching 10% of maxima and minima around the mean, for example with this line of code: h = patch([time time(end:-1:1) time(1)], [avg_min avg_max(end:-1:1) avg_min(1)], 'HandleVisibility', 'Off'); uistack(h, 'bottom'); where avg_min = avg - minima /10 and avg_max = avg + maxima /10 - chronjob ft_qualitycheck at Odin - chronjob should match with ft_qualitycheck 'exportname' - output .mat file has info, timelock, freq, summary and headpos structure - timecourses of headcoil distances - cleaned up code - updated help - timelimits: max 1 hour, 2nd plot when exceeding - triggers/events found - remove quantification sliders