No method matching curve_fit

[clemzm]

Hello ! I'm new to Julia and having trouble fitting a function of the form f(t,p)=p[1]+p[2]exp(-t/p[3]).
I tried to do this with curve_fit but I get the following error message:

f_exp(t::Float64,p::Vector{Float64})=p[1]+p[2]*exp(-t/p[3]);

p0=[1,1,0.5]
t=1:len

CurveFit.curve_fit(f_exp, t , P_TOD[1,:] , p0)

MethodError: no method matching curve_fit(::typeof(f_exp), ::UnitRange{Int64}, ::Vector{Float64}, ::Vector{Float64})
Closest candidates are:
curve_fit(::Type{Polynomial}, ::Any, ::Any, ::Any) at /home/heavybird/.julia/packages/CurveFit/JGc0o/src/linfit.jl:154
curve_fit(::Type{RationalPoly}, ::Any, ::Any, ::Any, ::Any) at /home/heavybird/.julia/packages/CurveFit/JGc0o/src/rationalfit.jl:92
curve_fit(::Type{RationalPoly}, ::Any, ::Any, ::Any, ::Any, ::Any) at /home/heavybird/.julia/packages/CurveFit/JGc0o/src/rationalfit.jl:92
...

Stacktrace:
[1] top-level scope
@ In[76]:6
[2] eval
@ ./boot.jl:360 [inlined]
[3] include_string(mapexpr::typeof(REPL.softscope), mod::Module, code::String, filename::String)
@ Base ./loading.jl:1116

[pkofod]

Sorry that I didn't get to you sooner, but you could use a vector type for x, so t = collect(1:len) would work