To try out this PR:

pip install https://12843-14579099-gh.circle-artifacts.com/0/dist/plotly-3.9.0%2B11.g6ec85ce5.tar.gz

Hi @jonmmease , I am getting an exception using this feature (I installed as per instruction above in my Conda environment running Python 3.7):

 print(plotly.__version__)
 fig = tools.make_subplots(
 rows=2,
 cols=2,
 print_grid=False,
 specs=[
 [{'secondary_y': True}, {'secondary_y': True}],
 [{'secondary_y': True}, {'secondary_y': True}]
 ]
 )

Outputs

3.9.0+11.g6ec85ce5
 File "/Users/marco/PycharmProjects/plotting/metrics_evolution_over_time/grouped.py", line 169, in ply_plotter
 [{'secondary_y': True}, {'secondary_y': True}]
 File "/Users/marco/anaconda3/envs/my_env/lib/python3.7/site-packages/plotly/tools.py", line 585, in make_subplots
 _check_keys_and_fill('specs', specs, SPEC_defaults)
 File "/Users/marco/anaconda3/envs/my_env/lib/python3.7/site-packages/plotly/tools.py", line 570, in _check_keys_and_fill
 _checks(arg_ii, defaults)
 File "/Users/marco/anaconda3/envs/my_env/lib/python3.7/site-packages/plotly/tools.py", line 563, in _checks
 "argument '{name}'".format(k=k, name=name))
Exception: Invalid key 'secondary_y' in keyword argument 'specs'

Looks like these lines in plotly.tools are missing a default for 'secondary_y':

 # Default spec key-values
 SPEC_defaults = dict(
 is_3d=False,
 colspan=1,
 rowspan=1,
 l=0.0,
 r=0.0,
 b=0.0,
 t=0.0
 # TODO add support for 'w' and 'h'
 )
 _check_keys_and_fill('specs', specs, SPEC_defaults)

Changing the above with a default works around the error.

 # Default spec key-values
 SPEC_defaults = dict(
 is_3d=False,
 colspan=1,
 rowspan=1,
 l=0.0,
 r=0.0,
 b=0.0,
 t=0.0,
 secondary_y=False
 # TODO add support for 'w' and 'h'
 )
 _check_keys_and_fill('specs', specs, SPEC_defaults)

However, it seems that my secondary_y are all ignored, as fig.print_grid() outputs

This is the format of your plot grid:
[ (1,1) x1,y1 ] [ (1,2) x2,y2 ]
[ (2,1) x3,y3 ] [ (2,2) x4,y4 ]

Did I do something wrong in the installation process?

I also tried pip install git+git://github.com/plotly/plotly.py.git@enh_secondary_y and got the same problems.

Hi @marcotama, thanks for giving it a try.

Do you have the from _plotly_future_ import v4_subplots line above the plotly import? This is needed to "turn on" the new subplots functionality. The version 4 make_subplot function should be in plotly.subplots.make_subplots. If that's not the issue, could you double check that there isn't a second version of plotly.py installed (maybe through conda?)

Thanks!

Hi @jonmmease , I feel a little stupid! I was missing that import line.
It works perfectly! Thanks!
Is this going to be part of 3.9.0? Will it need the _plotly_future_ line?

Hi @marcotama, great! Thanks for giving it a try. The current stable version is 3.9.0 so this will go into 3.10.0.

And yes, prior to version 4.0 you'll need this _plotly_future_ line to use this functionality. In version 4 this will become the default behavior. So the secondary_y functionality should not be considered stable until version 4, but help with testing is definitely appreciated so please open bug reports for any trouble you run into while using the _plotly_future_ flags. Thanks!