Gaurav Dhingra comments

Results 181 comments of


                                            Gaurav Dhingra

Add common test for classifiers reducing to less than two classes via sample weights during fit

I think we should raise `ValueError` in the first case, just as we did in the `GradientBoostingClassifier` (in PR mentioned in issue description). In the first case, I think it...

Add common test for classifiers reducing to less than two classes via sample weights during fit

I haven't tried to understand you code snippet. But I liked to point out that some of the classifier work ok with single class (even after sample_weight trimming), like `RandomForestClassifier`...

rrule for map

Using `git bisect` I can see that it was removed from system in commit: https://github.com/JuliaDiff/ChainRules.jl/commit/02e7857e34b5c01067a288262f69cfcb9fce069b

Some LinearAlgebra rrules

Hi @willtebbutt I would like to work on this issue. I will try to read up on it first and see how that goes, would that be ok?

What to do if `@thunk` is applied to a variable (rather than an expression) ?

So if I understand you correctly then: ```jl julia> function foo(x) dA = goo(x) + 1 return @thunk(dA) end foo (generic function with 1 method) julia> goo(x) = x^2 +...

What to do if `@thunk` is applied to a variable (rather than an expression) ?

Understood.

unthunk `getindex` and `iterate` on `Composite` objects

Currently something like: ```jl julia> r = Composite{Tuple{Float64,}}(a=(@thunk 2.0^2), b=(@thunk 2.0^3)) julia> collect(r) MethodError: Cannot `convert` an object of type Float64 to an object of type Thunk Closest candidates are:...

unthunk `getindex` and `iterate` on `Composite` objects

Though as expected `collect(Float64, r)` works just fine: ```jl julia> collect(Float64, r) 2-element Array{Float64,1}: 4.0 8.0 ``` I'll add the test case for it.

unthunk `getindex` and `iterate` on `Composite` objects

Hi @nickrobinson251 can you please review this :)

unthunk `getindex` and `iterate` on `Composite` objects

What about something like: ``` julia> a = Composite{Tuple{Thunk}}((@thunk 2.0^3), (@thunk 2.0^2)) ``` Is it supposed to give unthunked values on iteration?