ucsb-network-data-science-2016

UCSB IGERT Network Data Science Boot Camp (2016) materials

This is initially just my portion of the boot camp for setup and 2 hours of initial instruction.

Setup

Optional in brackets:

• Python: Anaconda distribution
  • IDE: Rodeo like RStudio
• [R, Rstudio]
• Git, Github. [GitKraken]

Git, Github

• git vs github
• Github Features
• Exercises:
  • setup Github account
  • Exercise for git: clone, add (json for class directory), commit, push, pull request
• [Github Pages, Markdown, Rmarkdown, presentation, website]

Python

Have notes, but do interactively

• Python console, basic calculator
• Jupyter Notebook
• everything is an object: dir
  • modules
• whitespace matters: loops, conditionals
• data types:

  • number (divide int vs float)

  • strings: real vs unicode

  • lists. list comprehension

  • dictionaries

  • Python Advanced: Graph Theory and Graphs in Python

  • Graph in NetworkX: networkx/graph.py at master · networkx/networkx:

    The Graph class uses a dict-of-dict-of-dict data structure. The outer dict (node_dict) holds adjacency information keyed by node. The next dict (adjlist_dict) represents the adjacency information and holds edge data keyed by neighbor. The inner dict (edge_attr_dict) represents the edge data and holds edge attribute values keyed by attribute names.

  • dates. import modules

    • Creating a graph — NetworkX 1.11 documentation

    • Graphs and networks¶ (github: simoninireland/cncp: complex networks, complex processes)

PANDAS

Tabular data (esp CSV)

• csv — CSV File Reading and Writing — examples

  • TODO: represent each row with own key,val vs each col has key: list of values

import csv

d = {} rdr = csv.reader(open('filename.csv', 'r')) d.keys = rdr.next() for row in rdr: k, v = row d[d.keys()] = v

-[pandas](http://pandas.pydata.org/pandas-docs/stable/) is well suited for "Tabular data with heterogeneously-typed columns, as in an SQL table or Excel spreadsheet"
- [Package overview — pandas 0.18.1 documentation](http://pandas.pydata.org/pandas-docs/stable/overview.html)
- [10 Minutes to pandas — pandas 0.18.1 documentation](http://pandas.pydata.org/pandas-docs/stable/10min.html)
- read csv (vs dic representation)
  ```python
  dic = pd.Series.from_csv(filename, names=cols, header=None).to_dict()

• data frames
• 12 Useful Pandas Techniques in Python for Data Manipulation

Projects

Both projects rely on creation of simpler networks from a dense raster for various applications:

• assessing spatial connectivity of habitats (Python)

  • extract TIN with cumulative distance away from patches for determining distances away
  • Keitt & Urban, Urban et al

• ship routing applications to avoid whale strikes (R)

  • increase density closer to shore

• how to make sparse networks from:

  • dens