Community Data Science Course (Spring 2023)/Week 3 lecture notes

Online Data Sets: An Important Question

Can you get bulk access to data?

Bad Signs:

• You must authenticate as a particular user in order to access data, and you can only see data for that user. (For example: you must log into Instagram's API as a particular user as per this link!)

Good signs:

• The organization owning the data wants everyone to access it. Like wikipedia or most government data.
• You may have to authenticate as a particular user, but you can access general data.
• For example: once you log into Reddit, you can get all posts about almost anything (Twitter API Docs)

Programming lecture outline

• Dictionaries!
  • Purpose (use dictionaries to store key/value pairs)
  • Initialization {}
  • Accessing elements
  • Adding elements
  • Changing elements
  • .values() and .values()
  • using in to look into dictionaries
  • using for loops to iterate over dictionaries (e.g., lets build a list of every letter in the alphabet using wordplay data)
  • A few notes about dictionaries:
    • A given key can only have one value, but multiple keys can have the same value.
    • Dictionaries do not guarantee ordering (although if you are using new versions of Python, order will be preserved).
• Additional loop control
  • break
  • continue
  • Note: These can be useful in combination to if statements and can also be super useful for debugging!
• Writing to files:
  • Using the open("whatever.tsv", "w") function
  • Using the with open() as my_file: statement
  • Writing to a file with print(file=my_file)
  • Writing a tab-separated value file using "\t" (make sure we leave a header!)
  • Now lets open it up and make a little graph
• Defining our own functions!
• A little bit on looking for help (if it hasn't come up already)
  • Looking at StackOverflow
  • Walking through the Python API documentation
  • Using a reference card or cheatsheet

Resources and Example Code

Initialization

>>> my_dict = {}
>>> my_dict
{}
>>> your_dict = {"Alice" : "chocolate", "Bob" : "strawberry", "Cara" : "mint chip"}
>>> your_dict
{'Bob': 'strawberry', 'Cara': 'mint chip', 'Alice': 'chocolate'}

Adding elements to a dictionary

>>> your_dict["Dora"] = "vanilla"
>>> your_dict
{'Bob': 'strawberry', 'Cara': 'mint chip', 'Dora': 'vanilla', 'Alice': 'chocolate'}

Accessing elements of a dictionary

>>> your_dict["Alice"]
'chocolate'
>>> your_dict.get("Alice")
'chocolate'

>>> your_dict["Eve"]
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
KeyError: 'Eve'
>>> "Eve" in your_dict
False
>>> "Alice" in your_dict
True
>>> your_dict.get("Eve")
>>> person = your_dict.get("Eve")
>>> print(person)
None
>>> print(type(person))
<type 'NoneType'>
>>> your_dict.get("Alice")
'chocolate'

Changing elements of a dictionary

>>> your_dict["Alice"] = "coconut"
>>> your_dict
{'Bob': 'strawberry', 'Cara': 'mint chip', 'Dora': 'vanilla', 'Alice': 'coconut'}

"Histograms"

Challenge: using wordplay example from last week, count the number of words that start with each letter.

This kind of problem is very common Data Science, and it is easy with a dictionary.

(note: I will post the solution after class)

For-loops and dictionaries

There are two common ways to iterate through dictionaries:

>>> ages = {'Tommy': 34, Heather: 30, 'Joanna': 20}
>>> for key in ages:
>>>     print(key + " is " + str(ages[key]) + " years old")

>>> for key, value in ages.items():
>>>     print(key + " is " + str(value) + " years old")