Problem using GLM

New to Julia
1

The following is throwing an error:

using DataFrames, GLM
data = DataFrame(X=1:1:length(event[1]), Y=event[1])
ols = lm(@formula(Y ~ X), data)

The reason appears to be to do with the type of data I define as X and Y:

typeof(X)   StepRange{Int64,Int64}

typeof(Y)   Vector{Any}

What am I missing?

2

first you don’t need the middle :1, and if the issue is about GLM doesn’t like StepRange then you can do X=collect(1:length(event[1])) instead in the second line to materialize an array from this iterator

3

The problem seems to be deeper:

image
image1239×176 13.5 KB

4

why does your array have a shape of 2? They need to be vectors ( Array{Float64,1})

5

Yea, what is typeof(event)? Your LHS variable needs to be a vector not a matrix. Somehow event[1] is producing a matrix for you. If it truly is a vector that is being misappropriated as a matrix then you can just wrap it in vec … vec(event[1])

1 Like
6

Just to add that you don’t have to collect the range when constructing the DataFrame, as this happens automatically:

julia> using GLM, DataFrames

julia> df = DataFrame(x = 1:10, y = 0.5 .* (1:10) .+ rand.())
10×2 DataFrame
│ Row │ x     │ y        │
│     │ Int64 │ Float64  │
├─────┼───────┼──────────┤
│ 1   │ 1     │ 0.874401 │
│ 2   │ 2     │ 1.07089  │
│ 3   │ 3     │ 2.08023  │
│ 4   │ 4     │ 2.89661  │
│ 5   │ 5     │ 3.22647  │
│ 6   │ 6     │ 3.0337   │
│ 7   │ 7     │ 3.81535  │
│ 8   │ 8     │ 4.19433  │
│ 9   │ 9     │ 4.69071  │
│ 10  │ 10    │ 5.7383   │

julia> df.x == collect(1:10)
true

julia> lm(@formula(y ~ x), df)
StatsModels.TableRegressionModel{LinearModel{GLM.LmResp{Array{Float64,1}},GLM.DensePredChol{Float64,LinearAlgebra.Cholesky{Float64,Array{Float64,2}}}},Array{Float64,2}}

y ~ 1 + x

Coefficients:
────────────────────────────────────────────────────────────────────────────
             Estimate  Std. Error   t value  Pr(>|t|)   Lower 95%  Upper 95%
────────────────────────────────────────────────────────────────────────────
(Intercept)  0.420504   0.220686    1.90544    0.0932  -0.0883982   0.929406
x            0.498472   0.0355667  14.0151     <1e-6    0.416455    0.580489
────────────────────────────────────────────────────────────────────────────
1 Like
7

My problem was solved by converting event[1] to a float:
convert(Array{Float64,1}, event[1])