feat: make ia_solve modular
#1348
Conversation
|
Codecov ReportAttention: Patch coverage is
❗ Your organization needs to install the Codecov GitHub app to enable full functionality. Additional details and impacted files@@ Coverage Diff @@
## master #1348 +/- ##
===========================================
+ Coverage 6.69% 78.93% +72.23%
===========================================
Files 47 50 +3
Lines 4618 4860 +242
===========================================
+ Hits 309 3836 +3527
+ Misses 4309 1024 -3285 ☔ View full report in Codecov by Sentry. |
| @eval fundamental_period(::typeof($fn)) = 360.0 | ||
| end | ||
|
|
||
| fundamental_period(::typeof(cospi)) = 2.0 | ||
|
|
||
| for fn in [tand, cotd] | ||
| @eval fundamental_period(::typeof($fn)) = 180.0 |
There was a problem hiding this comment.
is there a reason that these are floats? any downsides for making them ints? curious
There was a problem hiding this comment.
My reasoning was that due to how BasicSymbolic works, calling fundamental_period will always be dynamic dispatch but if all the return types are Float64, Julia should be able to infer that.
test/solver.jl
Outdated
| expr = sec(x ^ 2 + 4x + 4) ^ 3 - 3 | ||
| roots = ia_solve(expr, x) | ||
| @test length(roots) == 6 # 2 quadratic roots * 3 roots from cbrt(3) | ||
| roots = ia_solve(expr, x; complex_roots = false) | ||
| @test length(roots) == 2 |
There was a problem hiding this comment.
if you have the time, maybe also check that these roots' numerical values also holds by evaluating them and comparing them to the known answer from mathematica/matlab.
|
This is really good! i like it; if you get bored of your work in julia> Symbolics.ia_solve(-2a + c + 1 / c, c)
┌ Warning: This expression cannot be solved with the methods available to ia_solve. Try a numerical method instead.
└ @ Symbolics ~/code/julia/Symbolics.jl/src/solver/ia_main.jl:27
ERROR: MethodError: no method matching iterate(::Nothing)
|
|
@ChrisRackauckas this is good to merge |
This attempts to:
ia_solveto handle any function that implements the inverse interface instead of hardcoding a list.ia_solveto handle periodic functions that are notsin,cosortan.ia_solve(specifically, the way it handles periodic functions and complex roots)@n0rbed I would really appreciate your review on this