Documentation improvement suggestions

[vladdez]

I am reading this page and some details seems missing or ambigous.

• "Every row is an event. Note that :latency is commonly the timestamp in samples, whereas :onset would typically refer to seconds." - there is no :onset variable
• "# we have multi channel support" - what does this sentence means? Multiple channels are supported? Why is it important?
• "We define a formula that we want to apply to each point in time" - what does time point means here? to each epoch?
• "#0 as a dummy, we will combine with data later" - also not clear for me what does it mean. What combined with what? Why it is important to know that 0 is a dummy and what dummy means here?
• "Fit the UnfoldLinearModel." this model comes out of the blue. What is it, why we use it specifically?
• "As you can see, there is a lot going on" - what exactly is going on here?

[vladdez]

Additionally, for https://unfoldtoolbox.github.io/Unfold.jl/dev/HowTo/multiple_events/

• include(joinpath(dirname(pathof(Unfold)), "../test/test_utilities.jl") ) # to load data dat, evts = loadtestdata("test_case_4b"); - I suppose this way of loading data is a bit outdated
• "When dealing with overlapping data, it is often necessary to plot multiple events" - why it is necessary? Events - you mean experimental conditions?
• "The type column of table evts contains two conditions: EventA and EventB. By default, the toolbox will search for these conditions." - search to do what?
• bf1 = firbasis(τ = (-0.4, 0.8), sfreq = 50, name = "stimulusA") - why it is called stimulusA, not EventA? Confusing
• plot_erp(results; stderror = true, mapping = (; col = :group)) - wasn't able to recreate it because data is needed, but I suppose there should be two columns, while it isn't. Also, the title "fixef" could be renamed to smth more user-friendly.

[vladdez]

https://unfoldtoolbox.github.io/Unfold.jl/dev/HowTo/custom_solvers/

• I see the whole page Custom Solvers, but there is no explanation on what is a Solver in general
• "Robust solvers automatically account for outlier trials, but they come at a significant computational cost." - what is meant by "account"? Removing outliers?

[vladdez]

https://unfoldtoolbox.github.io/Unfold.jl/dev/HowTo/lmm_pvalues/

• Walds T-Test This method is easier to calculate but has limitations in accuracy and scope. It may also be less accurate due to estimation of degrees of freedom. - why DOFs hinders accuracy?
• in a section of methods comparison would be nice to write titles over each plot box

[vladdez]

https://unfoldtoolbox.github.io/Unfold.jl/dev/tutorials/lm_mu/

• "In some way, we are fitting a model with coefficients and then obtain the 'original' ERPs" - This sentence is obscure to me. Every model is fitted with coefficients, isn't it? What original ERP means here?
• "If a variable is not specified" - what kind of variable?
• "# Or we can split it up by condition and calculate all combinations automagically." - what is splitted?

[vladdez]

would be also nice to write a standart name and surname here:

[vladdez]

would be also nice to write a standart name and surname here:

this missed

[behinger]

ah, that is automatically done by the github, if someone provides their name publicly, fine. if not, fine as well