Share
This post is public.
I remember there was a search bar, we have accumulated about ~30 projects, and it's uncomfortable to scroll down or bring the web browser search option to find a specific project.
Some kind of organization options would be great, like folders (Metabase calls them "collections"), filtering by user, or something like that.
Some kind of organization options would be great, like folders (Metabase calls them "collections"), filtering by user, or something like that.
Share
This post is public.
I'm somewhat curious, is anyone using Deepnote for processes outside Analytics?
Right now, I'm optimizing some processes for Sales in the company I work for, things like batch pushing fresh data into our CRM or playing with our internal API to do "experiments."
I've found that it enables people with no code experience to execute custom code to improve their work-life. It's great, no setup is needed, and it's a perfect way to introduce them to code, especially if you're in a SaaS company.
Right now, I'm optimizing some processes for Sales in the company I work for, things like batch pushing fresh data into our CRM or playing with our internal API to do "experiments."
I've found that it enables people with no code experience to execute custom code to improve their work-life. It's great, no setup is needed, and it's a perfect way to introduce them to code, especially if you're in a SaaS company.
Share
This post is public.
Share
This post is public.
Not a showcase per se, but for a while now I've been trying to make H2O work with Deepnote, and the solution was pretty straight forward using the Custom Environment feature.
To give you a bit of context, H2O is a machine learning framework that helps you build models quickly using leading algorithms, it integrates AutoML features, and allows you to productionize your model using MOJO/POJO formats.
You can see an example of the project here: https://deepnote.com/project/37976cf6-7e17-4fad-aac2-1eafdecf8341
If you want your own H2O environment you can just duplicate my workspace by clicking on the Duplicate button, and if you're interested in how it's built then you can inspect the docker file by going to Environment > Dockerfile
Hope this helps anyone who's looking to marry DN with H2O π
To give you a bit of context, H2O is a machine learning framework that helps you build models quickly using leading algorithms, it integrates AutoML features, and allows you to productionize your model using MOJO/POJO formats.
You can see an example of the project here: https://deepnote.com/project/37976cf6-7e17-4fad-aac2-1eafdecf8341
If you want your own H2O environment you can just duplicate my workspace by clicking on the Duplicate button, and if you're interested in how it's built then you can inspect the docker file by going to Environment > Dockerfile
Hope this helps anyone who's looking to marry DN with H2O π
Share
This post is public.
I'm loving the SQL feature, I think it's a huge first step into making Deepnote a polyglot notebook. However, the feature is currently limited to executing a SQL query and returning the result in one single cell without the possibility of making further changes to that data frame.
My suggestion is to store the SQL output into a variable, this will allow us to further transform the data frame using Python and it would be greatly beneficial to the user's experience.
ο»Ώ
ο»Ώ
My suggestion is to store the SQL output into a variable, this will allow us to further transform the data frame using Python and it would be greatly beneficial to the user's experience.
ο»Ώ
Share
This post is public.
From a team management perspective it would make sense to pin notebooks at the top of the list. If I'm inviting a user to our platform I would want them to read an introductory notebook so that they can learn how to use DN, or maybe I want to pin topics that are very important right now and want them prioritized.
Share
This post is public.
As a user, I have to constantly repeat steps such as connecting to a DB, dumping CSV files, serializing models, plotting charts, and what-not. Therefore, I found myself going back to old notebooks to copy and paste snippets of code that I can reuse.
When working with teams or scaling data skills throughout an organization this can be more evident as you will find people asking how to do X thing.
Having a snippet gallery would help to mitigate this situation, make it more user-friendly to new users, and remove the necessity of turning on old notebooks to find code that you can reuse.
When working with teams or scaling data skills throughout an organization this can be more evident as you will find people asking how to do X thing.
Having a snippet gallery would help to mitigate this situation, make it more user-friendly to new users, and remove the necessity of turning on old notebooks to find code that you can reuse.