Change path to Python binary

[jannes-m]

Dear reticulate-developers,

thanks for developing such a fantastic package!

Is it possible to change the path to another Python binary after already having run py_config or py_run_string?
In my case I would like to point reticulate to the QGIS Python binary (which is a bit of a hazzle under Windows). This works, i.e. reticulate accepts the QGIS Python binary (mostly to be found under C:\OSGEO4~1\bin\python.exe).

However, before running any RQGIS function I would like to check if the correct Python binary is in use. Assuming, the user has not set up correctly the QGIS Python environment, running py_config might point to another Python binary such as Anaconda Python ("...\AppData\Local\CONTIN1\ANACON1\python.exe). In this case, I would like to change to the QGIS Python binary. I tried use_python but it didn't work. Or is there any other way to find out which Python binary reticulate would use at a specific moment but without actually setting it.

[jjallaire]

I agree that it would be good to have a "preview" of what the reticulate binding will be. I'll try to add that to the next version. In the meantime you can add the required = TRUE flag to use_python to fail with an error if it can't find the version specified.

[jannes-m]

Thanks for the answer! Until the "preview"-option becomes available, I will use a simple work-around.

I hope you don't mind another related question. Is it also possible to close (and then again to restart) the connection to Python? This would be rather useful. So far, I always restart R to achieve this.

[jjallaire]

No, you can't safely unload and reload the connection to Python (too many unreversable side effects in the Python interpreter)

…

On Thu, Mar 23, 2017 at 7:23 AM, Jannes Muenchow ***@***.***> wrote: Thanks for the answer! Until the "preview"-option becomes available, I will use a simple a work-around. I hope you don't mind another related question. Is it also possible to close (and then again to restart) the connection to Python? This would be rather useful. So far, I always restart R to achieve this. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#27 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAGXx08zbffF3kjHyUpopwJwe4hAZRhDks5rolY3gaJpZM4Mjmyc> .

[jjallaire]

Just exported the py_discover_config function here: 4118c57

[jannes-m]

Thanks for py_discover_config!

[mxposed]

@jjallaire could you please elaborate on why is it not possible to close established python session? I think it would be a nice feature to be able to re-initialize python with another binary (version or virtualenv) without restarting R session

[jjallaire]

You can't cleanly unload embedded Python and then load another version. There are too many side effects (e.g. static caches of objects) in embedded Python to allow for this. It is in-theory possible but in practice there are many bugs people have discovered (e..g https://bugs.python.org/issue17408) as most Python C code is not written with this scenario in mind.

…

On Tue, Jul 16, 2019 at 3:58 PM Nikolay Markov ***@***.***> wrote: @jjallaire <https://github.com/jjallaire> could you please elaborate on why is it not possible to close established python session? I think it would be a nice feature to be able to re-initialize python with another binary (version or virtualenv) without restarting R session — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#27?email_source=notifications&email_token=AAAZPR42YBG4J3JMWIQFTWTP7YR7HA5CNFSM4DEONSOKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD2B66YY#issuecomment-511962979>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAAZPR4GHB5FOBRGS6SVF2DP7YR7HANCNFSM4DEONSOA> .

[mxposed]

Thank you for the explanation. Is it theoretically possible to initialize python in a forked process and then kill one if there is a need for reload? But I guess the communication would become a lot more complex

[jjallaire]

Yes, if we initialize it in a forked process then we've lost the ability to R and Python to seamlessly share data / context in the foreground process (the R session).

…

On Wed, Jul 17, 2019 at 10:03 AM Nikolay Markov ***@***.***> wrote: Thank you for the explanation. Is it theoretically possible to initialize python in a forked process and then kill one if there is a need for reload? But I guess the communication would become a lot more complex — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#27?email_source=notifications&email_token=AAAZPR2N7OIIAX7O5355IGLP74RCXA5CNFSM4DEONSOKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD2EJRGQ#issuecomment-512268442>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAAZPRYDI25UGT3A2MM675DP74RCXANCNFSM4DEONSOA> .