Slickdeals Analytics with Pandas and Plotly

Slickdeals_Analytics.py

import pandas as pd
import plotly.express as px

# Parse Data
url="https://raw.githubusercontent.com/ChengzhiZhao/jupyter-notebooks/master/slickdeals_data_file.csv"
df=pd.read_csv(url,header=None,names=['datetime','store','title','item_url','deal_price','original_price', 'like', 'comments'])
df['datetime'] = pd.to_datetime(df['datetime'])

# Data Anlytics#
# Frequency
frequency_frontdeal = df.groupby(['title'])['title'].agg(['count']).reset_index(drop=False).sort_values(by=['count'],ascending=False)
fig = px.bar(frequency_frontdeal.head(20), x='title', y='count')
fig.show()

# Top Deal Trend By Day
airpod = df[df['title']=='Apple AirPods Pro w/ Wireless Charging Case'].sort_values(by=['like'])
fig = px.line(airpod, x='datetime', y='like')
fig.show()

# Why the trend looks like this?
airpod.groupby('store')['store'].count()
# store
# Google Shopping       96
# Verizon Wireless     473
# Name: store, dtype: int64

# With store
fig = px.line(airpod, x='datetime', y='like',color='store')
fig.show()

# Likes and Comments relationship
top_10_deals = frequency_frontdeal.head(10)
top10_df = df[df['title'].isin(top_10_deals['title'].tolist())]

transformed_top10_df_like = top10_df[['datetime','title','like']].rename(columns={'like':'count'})
transformed_top10_df_comments = top10_df[['datetime','title','comments']].rename(columns={'comments':'count'})
transformed_top10_df_like['type'] = 'like'
transformed_top10_df_comments['type'] = 'comments'

transformed_top10_df = pd.concat([transformed_top10_df_like, transformed_top10_df_comments])

fig = px.line(transformed_top10_df, x='datetime', y='count', color='type',facet_row="title")
fig.show()

# Animation
airpod['day'] = df['datetime'].dt.day
fig = px.line(airpod, x='datetime', y='like',color='store', animation_frame='day', animation_group='store')
fig.show()