Li Xiucheng

It's an indispensable feature. It's really surprised me when I found gorilla do not support emacs shortcuts in Mac OS X, as far as I know almost any place where...

> Does it work? All examples in this repository have been tested with PyTorch 0.4 and Julia 1.0. > Is it possible or reasonable efficient to compose arbitrary pytorch models...

> I notice there is no return statement at the end of forward method, is it so in Julia? Yes, all expressions have default values in Julia and the `return`...

Yes, that's correct. So we have to resort to the API like `index_select` by now.

[Tensor selection](https://github.com/boathit/JuliaTorch/blob/master/maskedAutoRegressiveFlow.jl#L51) and [tensor assignment](https://github.com/boathit/JuliaTorch/blob/master/maskedAutoRegressiveFlow.jl#L57).

It seems we cannot slice a tensor along multiple dimensions simultaneously so far.

There is no need to install. These examples only serve to illustrate how to use PyTorch in Julia by `PyCall.jl`.

That is easy. You can call `trip2seq` function to transform all trips in `porto.h5` to sequences, and save them into `trj.t`, just like [this line](https://github.com/boathit/t2vec/blob/master/experiment/utils.jl#L122).

You can you use either [`cell2gps`](https://github.com/boathit/t2vec/blob/master/preprocessing/SpatialRegionTools.jl#L139) to get the centroid gps of the cell or [`seq2trip`](https://github.com/boathit/t2vec/blob/master/preprocessing/SpatialRegionTools.jl#L293) to transform a sequence of cells into their gps locations.

It is a preserved interface at the time of its implementation just in case that we might want to change the `vocab_start` in the future. You can ignore it in...