Time-series Forecast ZenFile #35
Conversation
Used in the step operator
You can use this data and upload it under your project in GCP
lukyrasocha
changed the title
|
The spell check is failing on 'iam', but it is an abbreviation of Identity and Access Management. |
|
@strickvl any possibility to add iam to the spell check ignore? |
I would probably just ignore it for now. We're going to replace |
There was a problem hiding this comment.
Hi,
First of all, It's an amazing ZenFile which showcases the power of ZenML, I do have some changes requested, but in overall I think:-
- It would be good if you write good explanation of every step which you write in
README.md. - I think you haven't added doc strings in your code, so add comments in every step and pipelines.
- I think it would be nice if you can be standardised a bit with other ZenFiles. Like the structure of the project, I think you can organise this project in better way.
- I also feel
README.mdis bit empty, so useful explanations would be nice. - I think for setting up this project, I need to have poetry, but I think you should give a traditional option as well for setting up this project just like we did in other ZenFiles because that is bit common. I also think you can reference a link which says how one can install poetry and use poetry for installation.
If I think there can be any other changes, I will add other reviews for sure. If you are complete with current feedbacks, you can re request me for reviewing it again.
Amazing work :)
| X_train=np.ndarray, X_test=np.ndarray, y_train=np.ndarray, y_test=np.ndarray | ||
| ): | ||
| df = data.copy() | ||
| cardinal_directions = {'N': 0.0, |
There was a problem hiding this comment.
this looks a bit long, what about adding it in config file?
There was a problem hiding this comment.
I changed the code a bit so it looks better and added a description of how the feature engineering works. I find it more clear if the code stays together instead of using a config file in this case
Co-authored-by: Ayush Singh <81796368+ayush714@users.noreply.github.com>
|
Hey @ayush714, thank you so much for the feedback. I tried to change and adjust all the things you mentioned, but please let me know any additional feedback that you might have. Thanks! |
There was a problem hiding this comment.
It looks good from my side, I will wait for other peoples to approve it maybe @AlexejPenner and @strickvl.
|
Hi, I will run your whole code on my system and see everything work as per your readme instructions. I will be done by 14th June ( most probably ). |
There was a problem hiding this comment.
I again did a small review, I have a comment in README, I see that you are starting from very scratch which is good, what if I already have those things set up? I just need to create my stack, so in that case, it might get confused for other peoples who have things setup, so it would be nice to have a general note on the steps which starts from scratch.
I will be running your code once and will give more feedback.
Co-authored-by: Ayush Singh <81796368+ayush714@users.noreply.github.com>
Co-authored-by: Ayush Singh <81796368+ayush714@users.noreply.github.com>
Add a description of how to upload the data set to bigquery from CLI and also separated GCP steps and Zenml steps into two separate sections based on the latest comments
|
In the readme I now separated the steps into two sections - Steps on how to set up a GCP project and Steps on how to set up components for zenml stack. That should clear things up. I also added a description on how to upload the data set into BigQuery from CLI |
There was a problem hiding this comment.
I ran it on my own system and it seems to be working, I didn't found any such issues, It will be helpful if you can add a diagram/pipeline diagram in high level so it will be easier for peoples to understand what's going on!?
However I don't feel a strong requirement for that, if u can it's good otherwise I am approving it.
There was a problem hiding this comment.
I love this! Very well explained and a valuable showcase. Thank you so much for the amazing contribution @lukyrasocha !
|
Ok im merging it now. Amazing contribution and we'll be happy to release this tomorrow to the world! |
Predict electricity power generation from wind forecast using
ZenML:Also contains a thorough description on how to set up and configure a google cloud project together with a zenml stack.