Week 4: Function

---

Section 1: Assignment QA

---

Section 2: Data Type (Continued) & Function

G. Dictionary

• Dictionary is a list of key → value pairs. Keys are unique identifiers of elements. Values can be of different data type.
• The main usage of dictionary is to look up a value based its on key.
• How to create a dictionary?
  • Use curly braces {}, separating elements with comma.
  • Each element contains a unique key and a value, separated by colon.
  • Keys and values can be of any data type
  • Format: {key1:value1, key2:value2}
• How to look up a value for a given key?
  • Format: dictionary[key]

# Suppose we have three participants in our class: peter (id=1001), yuner (id=1002), benjamin (id=1003).
# Try generating a dictionary of participants' ids and their names.
student_dic={1001:"peter",1002:'yuner',1003:'benjamin'}

student_dic[1001]

# We can expand the dictionary to have more values about students, such as gender and age
student_dic={1001:["peter","M",23],1002:['yuner','F',32],1003:['benjamin','M',24]}

student_dic

We can use method .keys() to access all keys and .values() to access all values.

# Try .keys() and .values() methods.

# How many items are there in student_dic?

Exercise

Create two dictionaries of the top 5 movies on IMDB (https://www.imdb.com/chart/top/).
1.First dictionary: Use ranking as the key. Include movie name, rating and release year as the values.
2.Second dictionary: Use movie name as the key. Include ranking, rating, and release year as the values.

To add new items, you can simply use the following syntax, much similar to variable assignment:

dictionary[key] = value

#add one more movie to the dictionary

**Tip** Dictionary has a famous counterpart in JavaScript named JSON. Compared with previous five data types, dictionary operates at a higher level as it can represent not only the values but also the relationship between values. Moreover, it is highly human readable and indexing a dictionary by key is often the fastest way to search in Python.

H. DataFrame

• Dataframe is a data structure supported by the pandas module, which is equivalent to table in common sense.
• How to create a DataFrame?
  • Use Pandas' Function pd.DataFrame(data=...[,index=...,columns=...]). Arguments in square bracket are optional.
  • Import from other data source, like csv, excel or json. For example, pd.read_csv(file path [, indexcol=..., header=...]).
• Dataframe structure: rows and columns
  • columns are also known as Series
• Look up a value by index: dataframe.loc[index]
• Look up a value by relative location: dataframe.iloc[index]
• Access a single column: dataframe[column_name]
• Access multiple columns: dataframe[[column1, column2 ...]]

! pip3 install pandas

import pandas as pd
#We have two ways to create a DataFrame, i.e. progressive way and import way.
#1.1: Progressive way BY ROW
a=pd.DataFrame([[1001,'petter','M','HKU'],[1002,'yuner','F','HKBU'],[1003,'benjamin','M','CityU']])

a

#change headers
a.columns=['id','name','gender','affiliation']

a

a.index=['a','b','c']

a

a.columns

#1.2: Progressive way BY COLUMN
a=pd.DataFrame()
a['id']=[1001,1002,1003]
a['name']=['petter','yuner','benjamin']
a['gender']=['M','F','M']
a['affiliation']=['HK','HKBU','CityU']

#check out the index for each row
a.index

a

#assign a column to be index column
a=a.set_index('id')

a

#Retrieve values by index
a.loc[1001]

#Slice dataframe by index
a.loc[1001:1003]

#Retrieve values by relative location
a.iloc[0]

#Slice dataframe by relative location
a.iloc[:3]

#Access a single column/series
a['name']

#Access two columns/series
a[['name','gender']]

#Value comparison
a['gender']=='M'

#filter dataframe based on value comparison results
a[a['gender']=='M']

Download the data here: https://juniorworld.github.io/python-workshop/doc/COMM_journals.csv

#2: Easy way
#format: pd.read_csv(file path [, header=..., index_col=...])
b=pd.read_csv('COMM_journals.csv',header=0,index_col=0)

b

b=pd.read_csv('https://juniorworld.github.io/python-workshop/doc/COMM_journals.csv',header=0,index_col=0)

b

b.head()

b.tail()

Differences in Format/Synatx:

• format of method: object.method_name(), e.g. L.sort(), L.remove(1)
• format of function: function_name(), e.g. max(L), min(L)
• format of attribute: object.attribute_name, e.g. b.shape, b.index, b.columns

#check the dimensions of the dataframe with .shape attribute
b.shape

#min, max, mode, median, sum, var, std of Journal Impact Factor

We can use methods:

• Dataframe.sort_values(column_name[, ascending=True]) to sort dataframe by a column
• Series.value_counts([normalize=False])
• Dataframe.pivot_table(index=column1, [column=column2,] values=column3, aggfunc=function_name) to generate aggregate cross-tabulation about the dataframe
  • A combination of splitting dataframe, applying some function, and combining results

b.sort_values('Journal Impact Factor')

b.sort_values('Journal Impact Factor',ascending=False)

#save Dataframe to a csv file
b.to_csv("output_data.csv")

Download the stackoverflow developer survey data here: https://juniorworld.github.io/python-workshop/doc/stack-overflow-developer-survey-2022-first1000.csv

#read survey data into dataframe c
c = pd.read_csv("https://juniorworld.github.io/python-workshop/doc/stack-overflow-developer-survey-2022-first1000.csv")

c.head()

c.columns

c.shape

#break down dataframe by Gender, EdLevel, and Age
#.value_counts() & .value_counts(normalize=True)

Quiz 1

https://www.menti.com/alyoht9a49ac

Break

---

Section 3: Built-in functions

A. File I/O:

• Use open(path[,mode='r']) function.
  • Modes: read ('r') or write ('w') or both ('r+') or append ('a')
• Input: .readlines() method will extract all content in the file as a list of strings. One paragraph, one string.
• Output: .write(string) method write the given content to the file, from the beginning of file if mode 'w' is used or from the bottom of file if mode 'a' is used.
• Save and Close File: .close() method
• Tips:
  • Add line break \n
  • Remove line break or whitespaces at the beginning or the end of a string: method .strip()
  • Use slash / to denote directory

# Create a new file named "text.txt" in the current folder

#add three new lines, separated by line break mark \n

#close it

# Create a child folder test
# Create another file named "text2.txt" in the child folder

# Create a file named "text3.txt" in the parent folder

# grandparent folder?

#Open text.txt file

#read the existing lines

#print out the first line

#calculate the length (number of characters, including spaces) of first sentence

B. If/Else Statement

• If/Else Statement is used to test whether a condition is True. If statement is always followed by a question, a comparison condition, a logical operator, which will give you a boolen value of True or False. If True, do something. If False, do something else.
• Format: if logical_condition1 :... (else: ...)
• Example:

if a==1: print('yes') #Block A else: print('no') #Block B``` >

**Extra Knowledge** If/Else statement can be upgraded into a If/Elif/Else statement. Elif = Else + If

Practice: Realization of Decision Tree

#Use If/Elif/Else statement to realize this decision tree

#Demonstration
#Use If/Elif/Else statement to allocate a patient with records as below:
a={'new patient':False,'unpaid bill':False}