Rename SIRParticleFilter to BootstrapFilter #44
Conversation
|
@lassepe do you agree with this name change? @michael-lim do you have any thoughts since you have recently been thinking about particle filters? |
|
To my understanding, "SIR" describes the resampling process while "bootstrap" refers to SIR with a specific proposal distribution (transition prior probability). However, the wording is certainly used in various different ways. I have also found survey papers in which both terms are under interchangeably. |
| @@ -1,7 +1,7 @@ | |||
| name = "ParticleFilters" | |||
| uuid = "c8b314e2-9260-5cf8-ae76-3be7461ca6d0" | |||
| repo = "https://github.com/JuliaPOMDP/ParticleFilters.jl" | |||
| version = "0.5.1" | |||
| version = "0.5.2" | |||
There was a problem hiding this comment.
I think this should be 0.6.0 now since it breaks existing code.
There was a problem hiding this comment.
According to SemVer this would be okay since all major version 0 releases may be incompatible with eachother. But in Julia there is the convention:
While the semver specification says that all versions with a major version of 0 are incompatible with each other, we have made that choice that a version given as 0.a.b is considered compatible with 0.a.c if a != 0 and c >= b.
There was a problem hiding this comment.
Yeah, I actually looked this up (https://semver.org/#how-should-i-handle-deprecating-functionality), and I am still not 100% sure if I understand it correctly it. According to that section, it looks like what you're actually supposed to do is deprecate with a minor version release, and then remove in the next major version (so in this case deprecate in 0.5.2 and remove in 0.6.0).
Before julia 1.5, deprecation warnings were on by default, so there was a big performance penalty for deprecating things, but now they are off by default except in tests. So, I think 0.5.2 is actually correct.
|
Other than my comment about the version number this looks good to me. I would |
This fixes #22