include lower truncation for normal distrubution

[ReneSkukies]

Lower truncation might be useful to simulate specific experiments where there was/ is an exclusion criterion, artificially changing the distribution.

E.g. in our Oddball experiment, we set a lower RT threshold, throwing out every trial below.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (17f3fe8) 72.82% compared to head (1e02814) 69.93%.

❗ Current head 1e02814 differs from pull request most recent head 69902f8. Consider uploading reports for the commit 69902f8 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #57      +/-   ##
==========================================
- Coverage   72.82%   69.93%   -2.90%     
==========================================
  Files          10       10              
  Lines         357      316      -41     
==========================================
- Hits          260      221      -39     
+ Misses         97       95       -2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[behinger]

hey! great idea!

• Could you add a unit-test?
• Could you modify the tutorial to include this option (or at least mention it?)
• Could you elaborate whether this allows for upper & lower truncation at the same time?

[ReneSkukies]

Can do the first two.
To the last: It does! Didn't check with UnfoldSim, but with the Distribution package. If you have a truncated distribution type, the function just adds the other truncation as well. I.e. a Truncated(LogNormal{Float64}(μ=5.0, σ=1.0); upper=10.0) becomes Truncated(LogNormal{Float64}(μ=5.0, σ=1.0); lower=1.0, upper=10.0) if you apply the truncated function a second time (with a lower bound)

[github-actions]

[JuliaFormatter] _{reported by reviewdog 🐶}

Suggested change

	onset_lognormal = LogNormalOnset(; μ = 3, σ = 0.25, offset = 0, truncate_upper = nothing, truncate_lower = nothing);
	onset_lognormal = LogNormalOnset(;
	μ = 3,
	σ = 0.25,
	offset = 0,
	truncate_upper = nothing,
	truncate_lower = nothing,
	);

[ReneSkukies]

As this is for the docs only anyway I would omit

[ReneSkukies]

hey! great idea!

* Could you add a unit-test?

* Could you modify the tutorial to include this option (or at least mention it?)

* Could you elaborate whether this allows for upper & lower truncation at the same time?

Everything implemented now. I only mentioned the truncate lower in the docs because I didn't want to prolong the parameter plots even further. If you still want me to do that I can though.