diff --git a/src/FISTA.jl b/src/FISTA.jl
index 6b196cea..cdec7473 100644
--- a/src/FISTA.jl
+++ b/src/FISTA.jl
@@ -54,7 +54,6 @@ Solve the problem: X = arg min_x 1/2*|| Ax-b||² + λ*g(X) where:
 * `A`                       - system matrix
 * `b::Vector{T}`            - data vector (right-hand side)
 * `reg::Regularization`     - regularization object
-* (`AHA=nothing`)           - normal operator adjoint(A)*A
 * (`startVector=nothing`)   - start vector
 * (`iterations::Int64=50`)  - maximum number of iterations
 * (`ρ::Float64=1.0`)        - step size for gradient step
@@ -62,23 +61,7 @@ Solve the problem: X = arg min_x 1/2*|| Ax-b||² + λ*g(X) where:
 * (`relTol::Float64=1.e-4`) - relative tolerance for stopping criterion
 * (`solverInfo = nothing`)  - `solverInfo` object used to store convergence metrics
 """
-function fista(A,b::Vector{T}, reg::Regularization; AHA=nothing, kargs...) where T
-  if AHA==nothing
-    return fista1(A,b,reg;kargs...)
-  else
-    return fista2(A,b,reg;kargs...)
-  end
-end
-
-"""
-This funtion implements the fista algorithm.
-Solve the problem: X = arg min_x 1/2*|| Ax-b||² + λ*g(X) where:
-   x: variable (vector)
-   b: measured data
-   A: a general linear operator
-   g(X): a convex but not necessarily a smooth function
-"""
-function fista1(A, b::Vector{T}, reg::Regularization
+function fista(A, b::Vector{T}, reg::Regularization
                 ; startVector=nothing
                 , iterations::Int64=50
                 , ρ::Float64=1.0
@@ -128,49 +111,49 @@ end
 
 # alternative implementation allowing for an optimized AHA
 # does not contain a stopping condition
-function fista2(A, b::Vector{T}, reg::Regularization
-                ; AHA=nothing
-                , startVector=nothing
-                , iterations::Int64=50
-                , ρ::Float64=1.0
-                , t::Float64=1.0
-                , solverInfo = nothing
-                , kargs...) where T
-
-  if startVector == nothing
-    x = A' * b
-  else
-    x = startVector
-  end
-
-  # if AHA!=nothing
-    op = AHA
-  # else
-  #   op = A'*A
-  # end
-
-  β = A'*b
-
-  xᵒˡᵈ = copy(x)
-
-  solverInfo != nothing && storeInfo(solverInfo,A,b,x;xᵒˡᵈ=xᵒˡᵈ,reg=[reg])
-
-  costFunc = 0.5*norm(res)^2+norm(reg,x)
-
-  for l=1:iterations
-    xᵒˡᵈ[:] = x[:]
-
-    x[:] = x[:] - ρ*(op*x-β)
-
-    reg.prox!(x, ρ*reg.λ)
-
-    tᵒˡᵈ = t
-
-    t = (1. + sqrt(1. + 4. * tᵒˡᵈ^2)) / 2.
-    x[:] = x + (tᵒˡᵈ-1)/t*(x-xᵒˡᵈ)
-
-    solverInfo != nothing && storeInfo(solverInfo,A,b,x;xᵒˡᵈ=xᵒˡᵈ,reg=[reg])
-  end
-
-  return x
-end
+# function fista2(A, b::Vector{T}, reg::Regularization
+#                 ; AHA=nothing
+#                 , startVector=nothing
+#                 , iterations::Int64=50
+#                 , ρ::Float64=1.0
+#                 , t::Float64=1.0
+#                 , solverInfo = nothing
+#                 , kargs...) where T
+#
+#   if startVector == nothing
+#     x = A' * b
+#   else
+#     x = startVector
+#   end
+#
+#   # if AHA!=nothing
+#     op = AHA
+#   # else
+#   #   op = A'*A
+#   # end
+#
+#   β = A'*b
+#
+#   xᵒˡᵈ = copy(x)
+#
+#   solverInfo != nothing && storeInfo(solverInfo,A,b,x;xᵒˡᵈ=xᵒˡᵈ,reg=[reg])
+#
+#   costFunc = 0.5*norm(res)^2+norm(reg,x)
+#
+#   for l=1:iterations
+#     xᵒˡᵈ[:] = x[:]
+#
+#     x[:] = x[:] - ρ*(op*x-β)
+#
+#     reg.prox!(x, ρ*reg.λ)
+#
+#     tᵒˡᵈ = t
+#
+#     t = (1. + sqrt(1. + 4. * tᵒˡᵈ^2)) / 2.
+#     x[:] = x + (tᵒˡᵈ-1)/t*(x-xᵒˡᵈ)
+#
+#     solverInfo != nothing && storeInfo(solverInfo,A,b,x;xᵒˡᵈ=xᵒˡᵈ,reg=[reg])
+#   end
+#
+#   return x
+# end