automation of insights export to pdf from notebook
I have a Dataiku notebook in a flow that creates various visualizations. After the visualizations are generated, I manually publish and export them as a PDF. I would like to automate this entire process, from generating the visualizations to publishing and exporting the PDF.
Could someone guide me on how to automate this workflow within Dataiku? Are there any specific tools or methods within Dataiku for automating the notebook execution and PDF export process?
Answers
-
Alexandru Dataiker, Dataiku DSS Core Designer, Dataiku DSS ML Practitioner, Dataiku DSS Adv Designer, Registered Posts: 1,226 Dataiker
You can run "Execute the notebook" step in scenario →
Let us know if that helps. -
can you elaborate more on how to use this , thanks !!
-
Turribeach Dataiku DSS Core Designer, Neuron, Dataiku DSS Adv Designer, Registered, Neuron 2023 Posts: 2,123 Neuron
Can you please elaborate what part of the documentation you are not clear with?
-
i do not find
"Execute the notebook"
step in scenario
-
Turribeach Dataiku DSS Core Designer, Neuron, Dataiku DSS Adv Designer, Registered, Neuron 2023 Posts: 2,123 Neuron
The scenario step is called "Export notebook". Once you add the step you can then enable "Execute notebook". I believe it's an HTML export not a PDF export though, one for you to test. You can also attach the notebook in the https://doc.dataiku.com/dss/latest/scenarios/reporters.html#mail-reporter
-
after this how do i update the dashboard accordingly and export it ?
-
its a exploratory analysis which is done , so i create a dash board out of that code notebook usually so that all the graphs and ouputs are intact, how do i automate this ? and then later automate the export of the pdf
-
Turribeach Dataiku DSS Core Designer, Neuron, Dataiku DSS Adv Designer, Registered, Neuron 2023 Posts: 2,123 Neuron
So I have re-read your post and while you made it all about exporting a notebook I think this has misrepresented your true requirement. Your actual requirement is to automate the creation of some visualisations into a PDF. So let's focus on that instead. First of all let's understand that a Jupyter notebook is usually a "playground" environment to do data analysis and exploration. Once you have some code that works you should move it to a Python recipe. That's why there is even a Create Recipe button directly from the notebook. So the first step for your code will be to be converted to a Python recipe. Then you need to create the PDF file in a Dataiku Managed folder. Here is a post showing how to upload a file to a Managed Folder. Finally you will execute Python recipe by adding the managed folder in a Build job step in a Dataiku Scenario which can deal with the scheduling and distribution of the PDF. You can add a mail Scenario Reporter which can include an attachment from a Dataiku Managed folder. In this case the attachment will be the PDF your recipe generated.
-
Thank you, i just want to know how do you send automated emails from files in managed folder ?
-
Turribeach Dataiku DSS Core Designer, Neuron, Dataiku DSS Adv Designer, Registered, Neuron 2023 Posts: 2,123 Neuron
You can add a mail Scenario Reporter which can include an attachment from a Dataiku Managed folder. This shows how to add a dataset to a reporter:
But you can also add a file in a folder.