Maybe an issue: should convolve use x[Near(t-ll), :]?

[b-r-hamilton]

I'm trying to figure out why my solutions have significantly larger values at the first time step than other time steps. I'm not sure if this is an issue, or if it would solve my problem, but I'm wondering if it's appropriate to use Near while time indexing in convolve

BLUEs.jl/src/BLUEs.jl

Line 605 in cabf403

return sum([E[ii,:] ⋅ x[Near(t-ll),:] for (ii,ll) in enumerate(lags)])

For times t-ll < t[1], this will just access t[1]. So in my case, where the first year we have data for is 1470yr, the model matrix will assume that, when trying to account for the lag at 1470yr, that every preceding year has the same data as 1470yr (at least this is my understanding of what's happening here). Would it make more sense for this to only access times that exist?

[b-r-hamilton]

suggested fix - add a conditional that only computes the above line if t-ll <= x.dims[1][1]
I tried this out in a commit that's in the attached branch

[b-r-hamilton]

Okay, new update. The conditional should actually be t - ll >= t1 - 2*unit.(t1). The 2 is because, for our lagged M matrix, at t = 1yr, the values are 0, and we want to actually have values for the first row of our E matrix.

[b-r-hamilton]

Okay, I think this is my final opinion on this. Except I'm now wondering if this is moot because we shouldn't be subsampling M.
But, for what it's worth it, I think the right conditional is

    xtime = x.dims[1]
    return sum([E[At(ll),:] ⋅ x[At(t-ll),:] for (ii,ll) in enumerate(lags) if t-ll in xtime])