Python for Students of the Humanities and Social Sciences

name: inverse
layout: true
class: center, middle, inverse

---
# Python Programming 
## for Humanities and Social Science Students
### Making Networks
---

layout: false
# "Rules" of the class
**Ask questions** as we go along.

- **Remember**: There's always more than one way to do things!

- The goal here is not to be perfect but to get you up and running quickly.

- Remember keep your own data in mind as we go through these slides.
- Much of class time will be dedicated to working through your own research.

---
name: inverse
layout: true
class: center, middle, inverse
---
# Ready?

# Part 1: Making a network
---
name: how
layout: false

.left-column[
# We'll use Python to...
]
.right-column[
- Create a network
- Visualize a network
- Export a network (to Gephi)
- Analyze a network
<img src="pics/network.png" alt="Network image">
]

---
name: how
layout: false

.left-column[
# Python
## .orange[Networkx] 
]
.right-column[
- We'll use the **networkx** library

- But there are many others

- First, how to download a python library "package"

---
##Using **pip** to download python libraries
If you don't have **pip** use **easy_install** to install it!

Now, assuming pip is installed, run the following:
```remark-code
pip install networkx

# OR if you're on a mac
sudu -H pip install networkx
```
You should see the packages being installed on your computer!
]

---
#Let's look at a network!

- Easiest way: load a built-in network.
- Famous graph of relationships between wealthy Medici families

###Citations
- Breiger R. and Pattison P. (1986). Cumulated social roles: The duality of persons and their algebras. Social Networks, 8, 215-256.
- Kent D. (1978). The rise of the Medici: Faction in Florence, 1426-1434. Oxford: Oxford University Press.

```remark-code
import networkx as nx

g = nx.florentine_families_graph()
print "There are %d nodes" % len( g.nodes() )

>>> There are 15 nodes
```
---
# Now let's explore a little more.
### Documentation is here: https://networkx.readthedocs.io

```remark-code
import networkx as nx

g = nx.florentine_families_graph()  # get a graph object 
for n in g.nodes():                 # returns a list of nodes in 
    print n                         # print the node id

>>>
Strozzi
Tornabuoni
Medici
Albizzi
Guadagni
Pazzi
Acciaiuoli
Bischeri
Peruzzi
Ginori
Salviati
Castellani
Lamberteschi
Ridolfi
Barbadori
```
---
name: how
layout: false

# Now let's visualize this network
```remark-code
import networkx as nx
import matplotlib.pyplot as plt # MatPlot is a very powerful and complex lib!

g = nx.florentine_families_graph()

graph_pos = nx.shell_layout(g) # Define how to place the nodes

# Set the colours and text
nx.draw_networkx_nodes(g, graph_pos, node_size=1000, node_color='blue', alpha=0.3)
nx.draw_networkx_edges(g, graph_pos)
nx.draw_networkx_labels(g, graph_pos, font_size=12, font_family='sans-serif')

# Show graph
plt.show()
```
Results?
---
name: how
layout: false
.left-column[
### Medici Family Relationships
]
.right-column[
<img src="pics/medici_family.png" alt="Medici families">

<img src="pics/question1.png" align=left width=50 height=50>
.blue[Question: Which family is most important? ] 
]
---
name: how
layout: false
.left-column[
## NetworkX
### .orange[Exporting graphs]
]
.right-column[
- Networkx and matplotlib are great tools for visualizing small networks quickly.

- Larger graphs need dedicated tools

- **Gephi** is the best known graph visualization tool.

- http://gephi.org

]

---

# NetworkX
Exporting graphs
- Export Medici family data
- Gephi reads gexf files (XML text file)

```remark-code
nx.write_gexf( g, "medici.gexf" )
```

The GEXF looks like
<img src="pics/medici_gexf.png">

---
#NetworkX and Gephi
- Same graph file visualized with **Gephi**
- ForceAtlas layout
- Size of nodes is degree (number of edges)
<img src="pics/medici_graphi_layout.png">
---
#NetworkX and Gephi
Lessons: Visualization helps to tell the story
- Gephi is much easier to work with
- Much more clear which family is more important

<img src="pics/question1.png" align=left width=50 height=50>
.blue[Question: Which family is most important? ]

---
# Analyzing networks...
---

---

### Questions?
---