Introduction

Before you begin, review the overview documentation and log onto the DNAnexus Platform

New to JupyterLab?

If you have never used a JupyterLab notebook before, please view this information:

• Jupyter Notebook Documentation

• Try Jupyter

• Jupyter Architecture

• Content Community

Introduction

We can interact with the platform in several different ways and install software packages in these different environments depending on what we are wanting to use and how we want to use it. Here is what we are explaining in these sections:

Why JupyterLab?

Many Data Science Tasks are Interactive

• Notebook- based analysis

• Exploratory Data Analysis (EDA)

• Data Preprocessing/ Cleaning

• Implementing new Machine Learning/ Model

• Building Workflows

Requesting an Instance

Use Single DXJupyter Instance if:

• Work can be done on a single machine instance

• Main Use Cases:

  • Python/R

  • Image Processing

  • ML

  • Stata

Use Spark Cluster DXJupyter If:

• Working with very large datasets that will not fit in memory on a single instance

• Using Apollo and querying a large ingested dataset

• Need to use Spark based tools such as dxdata, HAIL or GLOW

JupyterLab with Use Cases

Spark JupyterLab with Use Cases

Starting an Instance

1. From Project List: Tools > JupyterLab

2. New JupyterLab button (top right)

3. Name your JupyterLab Environment

4. Snapshot (if you have one): single_jupyterlab.tar.gz

5. Project: Select Your Project

6. Choose your configuration

7. Choose your instance type

8. Choose your duration

9. Select your feature

10. Click Start Environment

Monitoring Your Instance

1. In the JupyterLab list, select the one you are wanting to monitor

2. On the right had side, info about the job will appear

   It will look like this:

1. You can also view this on the monitor tab by selecting in the info "View this Job in Monitor" (highlighted in gold above) or by using the Monitor Tab in the project space. It will look like this on the Monitor Tab

   

Running instances may take a while to load as the allocations become available.

1. Once it says "ready" select open

Resources

DXJupyterLab Quickstart Guide

Running DXJupyterLab

Full Documentation

To create a support ticket if there are technical issues:

1. Go to the Help header (same section where Projects and Tools are) inside the platform

2. Select "Contact Support"

3. Fill in the Subject and Message to submit a support ticket.