diff --git a/src/robustlinearmodel.jl b/src/robustlinearmodel.jl
index a45b13c..09a2733 100644
--- a/src/robustlinearmodel.jl
+++ b/src/robustlinearmodel.jl
@@ -98,14 +98,8 @@ function StatsAPI.fit(
     end
 
     # Make sure X and y have the same float eltype
-    T = promote_type(float(eltype(X)), float(eltype(y)))
-    if !(T <: AbstractFloat)
-        msg = "promoting X and y arrays to float types"
-        throw(TypeError(:fit, msg, Type{<:AbstractFloat}, T))
-    end
-    MT = AbstractMatrix{T}
-    VT = AbstractVector{T}
-    return fit(M, convert.(T, X)::MT, convert.(T, y)::VT, args...; kwargs...)
+    pX, py = promote_to_same_float(X, y)
+    return fit(M, pX, py, args...; kwargs...)
 end
 
 ## Convert from formula-data to modelmatrix-response calling form
@@ -125,14 +119,8 @@ function StatsAPI.fit(
     # Extract arrays from data using formula
     f, y, X, extra = modelframe(f, data, contrasts, dropmissing, M; wts=wts)
     # Call the `fit` method with arrays
-    T = promote_type(float(eltype(X)), float(eltype(y)))
-    if !(T <: AbstractFloat)
-        msg = "promoting X and y arrays to float types"
-        throw(TypeError(:fit, msg, Type{<:AbstractFloat}, T))
-    end
-    MT = AbstractMatrix{T}
-    VT = AbstractVector{T}
-    return fit(M, X::MT, y::VT, args...; wts=extra.wts, contrasts=contrasts, __formula=f, kwargs...)
+    pX, py = promote_to_same_float(X, y)
+    return fit(M, pX, py, args...; wts=extra.wts, contrasts=contrasts, __formula=f, kwargs...)
 end
 
 
diff --git a/src/tools.jl b/src/tools.jl
index cc99859..0356d02 100644
--- a/src/tools.jl
+++ b/src/tools.jl
@@ -4,6 +4,20 @@
 ##    Missing values
 ################################################
 
+function promote_to_same_float(
+    X::AbstractMatrix,
+    y::AbstractVector,
+)
+    T = promote_type(float(eltype(X)), float(eltype(y)))
+    if !(T <: AbstractFloat)
+        msg = "promoting X and y arrays to float types"
+        throw(TypeError(:fit, msg, Type{<:AbstractFloat}, T))
+    end
+    MT = AbstractMatrix{T}
+    VT = AbstractVector{T}
+    return convert.(T, X)::MT, convert.(T, y)::VT
+end
+
 _missing_omit(x::AbstractArray{T}) where {T} = copyto!(similar(x, nonmissingtype(T)), x)
 
 function StatsModels.missing_omit(X::AbstractMatrix, y::AbstractVector)