basis tests pass again

src/models/Rnl_splines.jl

@@ -133,9 +133,9 @@ end
 function rrule(::typeof(evaluate_batched), 
                basis::SplineRnlrzzBasis, 
                rs, zi, zjs, ps, st)
-   Rnl, st = evaluate_batched(basis, rs, zi, zjs, ps, st)
+   Rnl = evaluate_batched(basis, rs, zi, zjs, ps, st)
 
-   return (Rnl, st), 
+   return Rnl, 
          Δ -> (NoTangent(), NoTangent(), NoTangent(), NoTangent(), 
               NamedTuple(), NoTangent())
 end

src/models/ace.jl

@@ -611,23 +611,24 @@ end
 __vec(Rs::AbstractVector{SVector{3, T}}) where {T} = reinterpret(T, Rs)
 __svecs(Rsvec::AbstractVector{T}) where {T} = reinterpret(SVector{3, T}, Rsvec)
 
-function evaluate_basis_ed(model::ACEModel, 
+function evaluate_basis_ed_old(model::ACEModel, 
                            Rs::AbstractVector{SVector{3, T}}, Zs, Z0, 
                            ps, st) where {T}
 
    if length(Rs) == 0 
       B = zeros(T, length_basis(model))
       dB = zeros(SVector{3, T}, (0, length_basis(model)))
-   end
+   else
 
-   B = evaluate_basis(model, Rs, Zs, Z0, ps, st)
+      B = evaluate_basis(model, Rs, Zs, Z0, ps, st)
 
-   dB_vec = ForwardDiff.jacobian( 
-            _Rs -> evaluate_basis(model, __svecs(_Rs),  Zs, Z0, ps, st),
-            __vec(Rs))
-   dB1 = __svecs(collect(dB_vec')[:])
-   dB = collect( permutedims( reshape(dB1, length(Rs), length(B)), 
-                               (2, 1) ) )
+      dB_vec = ForwardDiff.jacobian( 
+               _Rs -> evaluate_basis(model, __svecs(_Rs),  Zs, Z0, ps, st),
+               __vec(Rs))
+      dB1 = __svecs(collect(dB_vec')[:])
+      dB = collect( permutedims( reshape(dB1, length(Rs), length(B)), 
+                                 (2, 1) ) )
+   end 
 
    return B, dB         
 end
@@ -655,7 +656,7 @@ end
 # ---------------------------------------------------------
 #  experimental pushforwards 
 
-function _evaluate_basis_ed(model::ACEModel, 
+function evaluate_basis_ed(model::ACEModel, 
                             Rs::AbstractVector{SVector{3, T}}, Zs, Z0, 
                             ps, st) where {T}
 
@@ -685,28 +686,37 @@ function _evaluate_basis_ed(model::ACEModel,
    end
 
    # pushfoward through the sparse tensor - this completes the MB jacobian. 
-   B, ∂B = _pfwd(model.tensor, Rnl, Ylm, ∂Rnl, ∂Ylm)
+   Bmb_i, ∂Bmb_i = _pfwd(model.tensor, Rnl, Ylm, ∂Rnl, ∂Ylm)
 
    # ------------------- 
    #  pair potential 
    if model.pairbasis != nothing    
       Rnl2, dRnl2 = @withalloc evaluate_ed_batched!(model.pairbasis, 
                                  rs, Z0, Zs, 
                                  ps.pairbasis, st.pairbasis)
-      Apair = sum(Rnl2, dims=1)[:]
-      ∂Apair = zeros(eltype(∂B), size(Rnl2, 2), size(Rnl2, 1))
+      B2_i = sum(Rnl2, dims=1)[:]
+      ∂B2_i = zeros(eltype(∂Bmb_i), size(Rnl2, 2), size(Rnl2, 1))
       for nl = 1:size(dRnl2, 2)
          for j = 1:size(dRnl2, 1)
-            ∂Apair[nl, j] = dRnl2[j, nl] * ∇rs[j]
+            ∂B2_i[nl, j] = dRnl2[j, nl] * ∇rs[j]
          end
       end
-
-      B = [ B; Apair ]
-      ∂B = [ ∂B; ∂Apair ]
-   end
+   else 
+      B2_i = zeros(eltype(Bmb_i), 0)
+      ∂B2_i = zeros(eltype(∂Bmb_i), 0, size(Bmb_i, 2))
+   end 
 
    # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~                     
    end  # @no_escape
 
+   TB = promote_type(eltype(Bmb_i), eltype(B2_i))
+   ∂TB = promote_type(eltype(∂Bmb_i), eltype(∂B2_i))
+   B = zeros(TB, length_basis(model))
+   B[get_basis_inds(model, Z0)] .= Bmb_i
+   B[get_pairbasis_inds(model, Z0)] .= B2_i
+   ∂B = zeros(∂TB, length_basis(model), size(∂Bmb_i, 2))
+   ∂B[get_basis_inds(model, Z0), :] .= ∂Bmb_i
+   ∂B[get_pairbasis_inds(model, Z0), :] .= ∂B2_i
+
    return B, ∂B 
 end

test/models/test_ace.jl

@@ -162,7 +162,9 @@ for ybasis in [:spherical, :solid]
       print_tf(@test Ei ≈ dot(B, θ))
 
       Ei, ∇Ei = M.evaluate_ed(model, Rs, Zs, z0, ps, st)
-      B, ∇B = M.evaluate_basis_ed(model, Rs, Zs, z0, ps, st)
+
+      B1, ∇B = M.evaluate_basis_ed(model, Rs, Zs, z0, ps, st)
+      print_tf(@test B ≈ B1)
       print_tf(@test ∇Ei ≈ sum(θ .* ∇B, dims=1)[:])   
    end
    println() 
@@ -208,6 +210,25 @@ for ybasis in [:spherical, :solid]
    end
    println() 
 
+## 
+
+   @info("After splinification check correctness of evaluate_basis_ed again")
+   for ntest = 1:10
+      local Nat, Rs, Zs, z0, Ei 
+      Nat = rand(8:16)
+      Rs, Zs, z0 = M.rand_atenv(model, Nat)
+      Us = randn(SVector{3, Float64}, Nat) / sqrt(Nat)
+      B, ∂B = M.evaluate_basis_ed(lin_ace, Rs, Zs, z0, ps, st)
+      B0 = M.evaluate_basis(lin_ace, Rs, Zs, z0, ps, st)
+      ∂B0 = M.jacobian_grad_params(lin_ace, Rs, Zs, z0, ps, st)[3]
+      print_tf(@test B ≈ B0)
+      print_tf(@test ∂B ≈ ∂B0)
+   end
+   println() 
+      # ∂E0 = ForwardDiff.derivative(
+      #       t -> M.evaluate_basis(lin_ace, Rs + t * Us, Zs, z0, ps, st), 
+      #       0.0)
+
 end
 
 ##