You could create a function which, given a symbol, generates the layer for
you:
```
getlayer(s) = layer(x=:x, y=s, Geom.line)
```
Next, you can use this function to get the layers you need:
```
plot(df, getlayer(:y1), getlayer(:y2), Guide.xlabel("k"))
# equivalently:
plot(df, layer(x=:x, y=:y1, Geom.line), layer(x=:x, y=:y2, Geom.line),
Guide.xlabel("k"))
```
If you have all the symbols you want to iterate over (check out the `names`
function...) you can use
```
plot(df, [getlayer(s) for s in symbols]..., Guide.xlabel("k"))
```
But a warning: this will probably be slow if you have many columns. It will
almost certainly be very slow with hundreds of them. (On the other hand, I
imagine the plot will be pretty hard to read with hundreds of lines...)
// T
On Monday, June 23, 2014 9:23:13 PM UTC+2, Leandro Seixas wrote:
>
> Hello,
>
> I am trying to do graphs with multiple layers with Gadfly. I have a
> dataframe with many columns, one "x" and a lot of "y"s, and I want to plot
> this in a single graph.
>
> For two layers, I have the following:
>
> plot(df, layer(x="x", y=df[2], Geom.line), layer(x="x", y=df[3],
> Geom.line), Guide.xlabel("k")),
>
>
> but if my dataframe has *dozens* or *hundreds* columns, what I should do
> to plot this in a single graph?
>
> Best,
>
> Leandro.
>
