Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/dogan-the-analyst/social_media_data_analysis
Data analysis with Python.
https://github.com/dogan-the-analyst/social_media_data_analysis
data-analysis jupyter-notebook matplotlib numpy pandas python
Last synced: about 1 month ago
JSON representation
Data analysis with Python.
- Host: GitHub
- URL: https://github.com/dogan-the-analyst/social_media_data_analysis
- Owner: dogan-the-analyst
- Created: 2024-12-25T14:02:02.000Z (about 2 months ago)
- Default Branch: main
- Last Pushed: 2025-01-01T17:03:26.000Z (about 1 month ago)
- Last Synced: 2025-01-01T18:18:23.474Z (about 1 month ago)
- Topics: data-analysis, jupyter-notebook, matplotlib, numpy, pandas, python
- Language: Jupyter Notebook
- Homepage:
- Size: 114 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Social Media Data Analysis
## Introduction
Social media has become a ubiquitous part of modern life, with platforms such as Instagram, Twitter, and Facebook serving as essential communication channels. Social media data sets are vast and complex, making analysis a challenging task for businesses and researchers alike. In this project, I explore a simulated social media, for example Tweets, data set to understand trends in likes across different categories.
## Project Scope
The objective of this project is to analyze tweets (or other social media data) and gain insights into user engagement. I will explore the data set using visualization techniques to understand the distribution of likes across different categories. Finally, I will analyze the data to draw conclusions about the most popular categories and the overall engagement on the platform.
## Importing Required Libraries
```python
import pandas as pd
from pandasql import sqldf
import numpy as np
import matplotlib.pyplot as plt
import random
```
## Creating Categories List, Data Size, and Random Data
```python
categories = ['Food', 'Travel', 'Fashion',
'Fitness', 'Music', 'Culture',
'Family', 'Health', 'Sports']
n = 1000
data_dict = {'Date': pd.date_range('2023-01-01', periods=n),
'Category': [random.choice(categories) for _ in range(n)],
'Likes': np.random.randint(0, 10000, size=n)}
```
## Transporting from Dictionary to DataFrame
```python
df = pd.DataFrame(data_dict)
df.tail()
```
## Exploratory Data Analysis (EDA)
### Checking Data Types
```python
df.dtypes
df.info()
```
### Dropping Duplicate Rows
```python
df.duplicated().sum()
```
- Great! Fortunately, there are no duplicate rows.
```python
df.drop_duplicates()
```
### Dropping Missing & Null Values
```python
df.isnull().sum()
```
- Again, no missing or null values.
```python
df = df.dropna()
df.count()
```
## Converting Fields
Even though the `Date` and `Likes` columns are in the correct format, I applied these formatting steps.
```python
df['Date'] = pd.to_datetime(df['Date'])
df['Likes'] = df['Likes'].astype('int64')
```
## Visualizing and Analyzing the Data
### Likes by Category
```python
likes_by_category = df.groupby('Category')['Likes'].sum()
likes_by_category.plot(kind="bar", figsize=(10, 5))
plt.title("Number of Likes by Category")
plt.ylabel("Number of Likes")
plt.xlabel("Category")
plt.xticks(rotation=45)
plt.show()
```
### Pie Chart of Likes Distribution
```python
likes_by_category = df.groupby('Category')['Likes'].sum()
plt.pie(likes_by_category, labels=likes_by_category.index)
plt.title("Distribution of Likes by Category")
plt.show()
```
### Mean of Likes
```python
mean_of_likes = df['Likes'].mean()
print("Mean of the 'Likes' category: {}".format(mean_of_likes))
```
### Number of Tweets by Category
```python
number_of_tweets_by_category = df.groupby('Category')['Category'].count()
number_of_tweets_by_category.plot(kind="bar", figsize=(10, 5), color='green')
plt.title("Number of Tweets by Category")
plt.ylabel("Number of Tweets")
plt.xlabel("Category")
plt.xticks(rotation=45)
plt.show()
```
### Pie Chart of Tweets Distribution
```python
number_of_tweets_by_category = df.groupby('Category')['Category'].count()
plt.pie(number_of_tweets_by_category, labels=number_of_tweets_by_category.index)
plt.title("Distribution of Tweets by Category")
plt.show()
```
### Most Popular Category
```python
query = """
SELECT
Category,
COUNT(*) AS Frequency
FROM
df
GROUP BY
Category
ORDER BY
Frequency DESC
"""
result = sqldf(query)
print(result)
```
- The result shows that `Sports` is in first place.