/dsc-network-community-detection-lab-dc-ds-010620

Primary LanguageJupyter NotebookOtherNOASSERTION

Community Detection - Lab

Introduction

In this lab, you'll once again work to cluster a network dataset. This time, you'll be investigating a Meetup dataset for events occurring in Nashville.

Objectives

You will be able to:

  • Make visualizations of clusters and gain insights about how the clusters have formed
  • Cluster a real-life social network dataset to detect communities within the network

Load the Dataset

To get started, load the dataset 'nashville-meetup/group-edges.csv' as a pandas Dataframe.

# Your code here

groups = None

To add some descriptive data, import the file 'nashville-meetup/meta-groups.csv'.

# Your code here
groups_meta = None

Transform to a Network Representation

Take the Pandas DataFrame and transform it into a graph representation via NetworkX. Make nodes the names of the various groups within the Nashville area. The edges should be the weights between these groups.

# Your code here

Visualize the Network

# Your code here

Refine the Visual

As you should see, the initial visualization is a globular mess! Refine the visualization to better picture the center core of the network.

# Your code here

Cluster the Network

Now, that the dataset is a little more manageable, try clustering the remaining group nodes.

# Your code here

Determine An Optimal Clustering Schema

Finally, determine a final clustering organization.

# Your code here

Visualize the Clusters

Analyze the output of your clustering schema. Do any clusters of groups stand out to you?

# Your code here

Summary

In this lab, you got another chance to practice your network clustering skills, investigating a meetup dataset of events and groups surrounding the Nashville area. If you're up for the challenge, feel free to continue investigating the dataset and trying to cluster users or events!