7:00 – 8:00 Research

• Information Wars: A Window into the Alternative Media Ecosystem
  • By Kate Starbird, (Scholar)
    Professor, Human Centered Design & Engineering, University of Washington
    Crisis Informatics, Crowdsourcing, Crowdwork, CSCW, Human Computer Interaction
• Started reviewing the Tumblr trolling paper. Set up a Tumblr account

8:30 – 5:00 BRC

• Heath was able to upgrade to Python 3.5.2
  • Ran array_thoughts. Numbers are better than my laptop
  • Attempting just_dbscan: Some hiccups due to compiling from sources. (No module named _bz7). Stalled? Sent many links.
  • Success! Heath installed a binary Python rather than compiling from sources. A little faster than my laptop. No GPUs, CPU, not memory bound.
• Continuing my tour of the SciPy Lecture Notes
• Figuring out what a matplotlib backend is
• Looks like there are multiple ways to serve graphics: http://matplotlib.org/faq/howto_faq.html#howto-webapp
• More on typing Python
• Class creation, inheritance and superclass overloading, with type hints:

  class Student(object):
      name = 'noName'
      age = -1
      major = 'unset'
  
      def __init__(self, name: str):
          self.name = name
  
      def set_age(self, age: int):
          self.age = age
  
      def set_major(self, major: str):
          self.major = major
  
      def to_string(self) -> str:
          return "name = {0}\nage = {1}\nmajor = {2}"\
              .format(self.name, self.age, self.major)
  
  
  class MasterStudent(Student):
      internship = 'mandatory, from March to June'
  
      def to_string(self) -> str:
          return "{0}\ninternship = {1}"\
              .format(Student.to_string(self), self.internship)
  
  
  anna = MasterStudent('anna')
  print(anna.to_string())

• Finished the Python part, Numpy next
• Figured out how to to get a matrix shape, (again, with type hints):

  import numpy as np
  
  
  def set_array_sequence(mat: np.ndarray):
      for i in range(mat.shape[0]):
          for j in range(mat.shape[1]):
              mat[i, j] = i * 10 + j
  
  
  a = np.zeros([10, 3])
  set_array_sequence(a)
  print(a.shape)
  print(a)