topoplot interpretation label

[vladdez]

We asked in a survey about time interpretation on topoplot and topoplot timeseries´, but don't have any ways to put that on a plot yet. Should be added.

[behinger]

these labels already exist

squrae => timepoint included, round timepoint excluded, according to the accepted notation for ranges

Did you have something else in mind?

[vladdez]

Hm, OK, but do we have a way of switching between the modes? There should be two modes: time point or average of time interval. Ideally it should be done by an argument.

[behinger]

mh that feature is not yet implemented. it is very easy to implement though. You could just use the following combinefun :

eeg_topoplotseries(...,combinefun = x->x[end÷2])

it needs a dedicated argument though, because this is a bit opaque ;-) What it does is instead of averaging (mean is default) it selects the middle element (if this is what one wants to specfy, and not the first sample) - what needs fixing further is the labels

[vladdez]

selects the middle element - is it the median?

will eeg_topoplotseries(...,combinefun = median) five the same result?

[vladdez]

and of course it should show one number, not the interval

[vladdez]

also we need such option of a single topoplot.

[behinger]

• no, selecting the middle value is not the same as the median. Just try it :-)
• exactly, it should show one number which is not implemented, I meant that with "fixing the labels"
• I dont understand why we would need that for a single topoplot. If you want to do that, you can calculate the mean / select one sample yourself

[vladdez]

1. I tried and see no difference visually (and theoretically tbh)
2. ok
3. for sake of reproducibility! looking at a single topoplot I want to know what it shows me: a single time point (and which one exactly), mean over all time points (or some time interval), what is the time interval used (sec or msec)? If we are advocating for label specification everywhere we should be strict ourselves.

[behinger]

between median, mean and selecting the middle value?

3. I cant think of a good interface, what exactly do you propose?

[vladdez]

basically the same as in a topoplot series: a title with a single number or time interval under the topoplot

[vladdez]

[behinger]

2. no, the result is clear, but the interface, how you call plot_topoplot differently.
3. for mean, median and middle value the result might look similar in smooth data, but it is extremely unlikely that they are the same

[vladdez]

1. why should we call it differently? I think it should be default feature. How to implement it? That why I am asking you ^^
2. the difference between mean and median is clear. The difference between median and middle is not clear, they are synonyms, isn't it? Like "the median is the middle value when a data set is ordered from least to greatest".

[behinger]

1. missunderstanding, how should we call the plot_topoplot signature differently, e.g. also offer combinefun?
2. exactly, the median is fo the sorted/ordered vector, which is not the case for choosing the midlevalue (a single sample)

[vladdez]

1. actually that is a hard question. A partial solution could be just to enable opportunity to manually specify xlabel.
2. okay