Fix/gradle

[guillaumeraille]

• change workingdir for notebook in CI to darts root
• update every notebook to use that path
• update CI coverage run to not include tests and __init__.py

Summary

Other Information

[Kostiiii]

In docker after you go to examples directory and open notebook the notebook current directory is examples. So I feel like the checks and workdir of it should stay in examples.

[guillaumeraille]

In docker after you go to examples directory and open notebook the notebook current directory is examples. So I feel like the checks and workdir of it should stay in examples.

I am not sure to follow, you mean when you use the docker image and run jupyter notebook from within the examples folder within the docker container ?

[Kostiiii]

In docker after you go to examples directory and open notebook the notebook current directory is examples. So I feel like the checks and workdir of it should stay in examples.

I am not sure to follow, you mean when you use the docker image and run jupyter notebook from within the examples folder within the docker container ?

I run ./gradlew docker && ./gradlew dockerRun then open link to jupyter, then from root folder I go to examples folder and open some notebook. Running %pwd inside it shows that workdir is set to examples.

[guillaumeraille]

In docker after you go to examples directory and open notebook the notebook current directory is examples. So I feel like the checks and workdir of it should stay in examples.

I am not sure to follow, you mean when you use the docker image and run jupyter notebook from within the examples folder within the docker container ?

I run ./gradlew docker && ./gradlew dockerRun then open link to jupyter, then from root folder I go to examples folder and open some notebook. Running %pwd inside it shows that workdir is set to examples.

I was trying to find a fix for it without success.
I also tried to change the path to use the filename directory in VSCode but the problem is that then it can't find darts in the path anymore when the path is set to examples. I believe living the working dir in the root folder is better as we can change the code and instantly inspect changes in cells calling the corresponding code

[guillaumeraille]

to add some clarification here is my problem @Kostiiii:

because of this path used inside the jupyter notebook in docker, when you actually make a change you have to build the package again to see some changes reflected in the notebook

[Kostiiii]

to add some clarification here is my problem @Kostiiii:

because of this path used inside the jupyter notebook in docker, when you actually make a change you have to build the package again to see some changes reflected in the notebook

Okey, now I understand what is your concern. Hmm I always though that notebooks are supposed to use installed version of darts not local code.

[guillaumeraille]

to add some clarification here is my problem @Kostiiii:

because of this path used inside the jupyter notebook in docker, when you actually make a change you have to build the package again to see some changes reflected in the notebook

Okey, now I understand what is your concern. Hmm I always though that notebooks are supposed to use installed version of darts not local code.

if that's what we want then we can discard this changes and I will create my own notebooks locally :)

[TheMP]

to add some clarification here is my problem @Kostiiii:

because of this path used inside the jupyter notebook in docker, when you actually make a change you have to build the package again to see some changes reflected in the notebook

Okey, now I understand what is your concern. Hmm I always though that notebooks are supposed to use installed version of darts not local code.

if that's what we want then we can discard this changes and I will create my own notebooks locally :)

@guillaumeraille Bit late to the party but would simply adding darts to the path in notebook help? https://www.johnny-lin.com/cdat_tips/tips_pylang/path.html?

[guillaumeraille]

to add some clarification here is my problem @Kostiiii:

because of this path used inside the jupyter notebook in docker, when you actually make a change you have to build the package again to see some changes reflected in the notebook

Okey, now I understand what is your concern. Hmm I always though that notebooks are supposed to use installed version of darts not local code.

if that's what we want then we can discard this changes and I will create my own notebooks locally :)

@guillaumeraille Bit late to the party but would simply adding darts to the path in notebook help? https://www.johnny-lin.com/cdat_tips/tips_pylang/path.html?

it would but then it would mean adding a code cell that should be run only locally. Actually, I think even adding something to the path that is not used would also work so running it in docker would be useless but fine. I can make the changes and try thanks

[TheMP]

Can't come up with anything better ATM :)