Easier way to make anonymous functions

[pdeffebach]

I'm imagining somehtin like @aside but which takes in _ and returns whatever the expression returns.

@chain 1 begin 
    @anon begin 
        _ + 2 * 1000
     end
     mean()
end

I've been thinking of doing an @anon macro in DataFramesMeta, since the infrastructure is all there to do it with Symbols. However this would need to be different, special-casing _.

[jkrumbiegel]

Can you give a more realistic example, what you wrote here works already without @anon around it

[pdeffebach]

Yeah, I'm thinking intermediate variables as well

@chain 1 begin 
    @anon begin 
        x = 500 
        z = 78
        _ + x * z
    end
end

[jtrakk]

Is that the same as this?

@chain 1 begin 
    y-> begin 
        x = 500 
        z = 78
       y + x * z
    end
end

[pdeffebach]

hmmm yes it is. That's definitely easier.

Though I guess that solution incurs a compilation cost that could be avoided if we let things happen in the normal scope. (This goes against my @anon name)

It also enables the user to leave the world of _, which complicates things.

[jkrumbiegel]

@chain 1 begin 
    begin 
        x = 500 
        z = 78
        _ + x * z
    end
end

[pdeffebach]

Oh no kidding. I thought I tried that. Excellent!