I saw - and happily tested - the new possibility of having inline
figures in the ipython notebooks.
Unfortunately (for my typical usage, in which I both save figures and
show them in the notebook, which is easy with matplotlib and networkx),
it seems like setting the "inline" argument makes the "ouput" argument
irrelevant.
I didn't realize anyone would actually be doing this, but I see no
reason not to. I've changed it now in the git version, so that the
output is no longer ignored with inline == True.
> I saw - and happily tested - the new possibility of having inline
> figures in the ipython notebooks.
>
> Unfortunately (for my typical usage, in which I both save figures and
> show them in the notebook, which is easy with matplotlib and networkx),
> it seems like setting the "inline" argument makes the "ouput" argument
> irrelevant.
>
> Am I wrong? Is this an intended behaviour?
I didn't realize anyone would actually be doing this, but I see no
reason not to. I've changed it now in the git version, so that the
output is no longer ignored with inline == True.
There was only a small problem for my typical workflow: I usually
produce figures in pdf, which however raises an error if I set
"inline=True", because it cannot be embedded in the webpage.
Another minor annoyance that I found with the inline ability is that it
is different from matplotlib and networkx in the fact that it must be
run as last.
That is, if you run a cell with the following content:
import networkx
g = networkx.random_graphs.complete_graph(5)
networkx.draw(g)
print "ciao"
... in the output you see both "ciao" and the network. If instead you
try
from graph_tool.all import *
g = complete_graph(5)
graph_draw(g, inline=True)
print("ciao")
... in the output you will only see "ciao".
I'm not expert in IPython internals, but I guess it should be simple to
solve.
There was only a small problem for my typical workflow: I usually
produce figures in pdf, which however raises an error if I set
"inline=True", because it cannot be embedded in the webpage.
Another minor annoyance that I found with the inline ability is that it
is different from matplotlib and networkx in the fact that it must be
run as last.
That is, if you run a cell with the following content:
import networkx
g = networkx.random_graphs.complete_graph(5)
networkx.draw(g)
print "ciao"
... in the output you see both "ciao" and the network. If instead you
try
from graph_tool.all import *
g = complete_graph(5)
graph_draw(g, inline=True)
print("ciao")
... in the output you will only see "ciao".
I'm not expert in IPython internals, but I guess it should be simple to
solve.
Yes, the image should be simply displayed instead of returned. I have
fixed this now in git.