Using models with data

[PharmCat]

Is it possible to use models with survival data and get likelihood estimates?

[alecloudenback]

For sure! I will try to create an example this weekend by combining this package with, e.g, LsqFit.jl

[PharmCat]

For sure! I will try to create an example this weekend by combining this package with, e.g, LsqFit.jl

Hi! I made an example, is my approach correct? (or maybe linearization need)

import LsqFit
using MortalityTables, Plots

data = [
1 0.99
2 0.98
3 0.95
4 0.9
5 0.8
6 0.65
7 0.5
8 0.38
9 0.25
10 0.2
11 0.1
12 0.05
13 0.02
14 0.01]

plot(data[:,1], data[:,2])
@. model(x, p) = survival(MortalityTables.Weibull(;m = p[1],σ = p[2]), x)
mfit = LsqFit.curve_fit(model, data[:,1], data[:,2], [1.0, 1.0])
plot!(data[:,1], model(data[:,1], mfit.param))

[alecloudenback]

Looks good to me!

[alecloudenback]

@PharmCat are you okay if I close this issue? Also, do you mind if I use this example in the documentation or the JuliaActuary.org website?

[PharmCat]

Also, do you mind if I use this example in the documentation or the JuliaActuary.org website?

Hi! I made some examples here: https://gist.github.com/PharmCat/fe1e64a2241aa35d3a3fa925989c3784 Of course, if you want, you can use it :) One thing... for data with censored values I used Weibull from Distributions for ML estimate (#3). Is it possible to make something like this with MortalityTables (Because MortalityTables includes more models)?

are you okay if I close this issue?

OK)

[alecloudenback]

One thing... for data with censored values I used Weibull from Distributions for ML estimate (#3). Is it possible to make something like this with MortalityTables (Because MortalityTables includes more models)?

Sorry, can you clarify what you mean here? I'm not sure that the link is pointing to the right issue.