Project "Find the Postcode"

By Netnito1010 - netnito1010@gmail.com

Intro

Consider a situation that we have gotten a data of addresses of companies. We want to send a mail about our new product to each of the company in the list. However, the addresses we have doesn't contain postcode.

The aim of this project is to teach you how to use basic pandas to extract 'Province', 'District', and 'Tambon' from a giving address and use that information to searching for the postcode in the pre-created database.

---

Table of Contents

• Project "Find the Postcode"
  • What we have and what we need before we begin this project
  • Project's Outline
  • Project's Step-by-Step Outline
    • 0 Setting up
    • 1 Import data
    • 2 Check the addresses data
    • 3 Have fun with the postcode data
    • 4 Have fun with the addresses data
      • 4.1 Clean the address

---

What we have and what we need before we begin this project

nothing. JK. I have provided you the Adresses data which is called SET_DATA.csv and the pre-created database of the postcode called ProCheck.csv.

Project's Outline

This is the pseudo code of the project

Project's Step-by-Step Outline

This is different from the main outline since we will explore the deeper detail of each step and might take some detours to see how each function works

Let's begin

0 Setting up

First check that your terminal is in the correct path (the path which you have your main python file and the data)

And don't for get that we are going to use the package pandas (as pd)

1 Import data

Q1: Import the data

Hint1: Now you should be able to import the data. Let's use the function pd.read_csv() (from pandas) to import the data into a variable call df and df_postcode for SET_Data.csv and ProCheck.csv respectively.

Hint2: Don't forget to specific the path of the data

Q2: Display the header useing head()

Hint2: google "head pandas"

2 Check the addresses data

Q1: Count the number of rows in the df data.

Hint: google "count row pandas"

Hint2: Don't for get to print it to see the results

Note: There many ways to count the rows and with some specific way it can slightly improve your performance.

3 Have fun with the postcode data

Q1: Count the number of rows in the df_postcode data

Q2: Count the number of unique province in the df_postcode data

Hint: First, you need to select the province column which called "Prov". Next, we count the unique value in that column - try google "count unique values pandas"

Q3: Display the data that Prov is "กรุงเทพมหานคร"

Hint: Use something like df[df["Prov"=="กรุงเทพมหานคร"]] and display it using print() or display() to show the dat

Q4: Display the data that Prov is "สมุทรปราการ" and Dist is "พระประแดง"

Hint: Use the similar idea as the provious question

Q5: Print the data that Prov is 'กรุงเทพมหานคร', Dist is 'พระนคร', and Tamb = 'วัดสามพระยา'

Hint: Use the similar idea as the provious question

Q6: Show the post code of when Prov is 'กรุงเทพมหานคร', Dist is 'พระนคร', and Tamb = 'วัดสามพระยา'

Hint: Try using iloc or at funtion to the first row of the table. Then select the PostCode column and print the value.

Hint2: try google 'iloc pandas' and 'print specific cell pandas'

Note: Without using iloc or at the data we get is dataFrame opject which is not the value in the cell. If you want to use just the value in that cell we need to use iloc or at.

4 Have fun with the addresses data

Here we will deal with the first few rows of the data then we use for loop to apply the idea to the whole data.

Q1: Store the column 'Location' of the data in

Q2: Store the first few row of data in df_location_head

Hint: You might need to use something like df.head()

Q3: Display the data and check the pattern

One might notice that the Province is always the last word in the address. The seccond last is District and the third last is Tambon. Therefore, we will use this information to extract the data we need.

4.1 Clean the address

Before we split the data we need to clean it first.

Task1: Replace 'กทม.' with 'กรุงเทพมหานคร'

Later, we will compare the province to the database and in the database Bangkok is refered as 'กรุงเทพมหานคร'.

Task2: Remove 'อำเภอ', 'ตำบล', 'อ.', and 'ต.'

Again, in the database we don't have the word 'อำเภอ', 'ตำบล', 'อ.', and 'ต.', therefore we don't need it.

Task3: Split the data and store the province, district, and tambon in a new database called df2