Skip to content

Latest commit

 

History

History
58 lines (45 loc) · 1.92 KB

File metadata and controls

58 lines (45 loc) · 1.92 KB

T5_Capital_Bikeshare

This is a data set of Capital Bikeshare users on an hourly basis across Washington D.C. This data was gathered from 2011 - 2012 and it is created by Data Society.

  • Link: data.world.
  • Or you can download it from the attached file.

Business Problem:

Predict the total users that depends on time.

Dataset Description:

This data has 17,379 rows, 14 columns.

Num Command Description
1 Date The date of the day
2 Season Seasons of the year -> 1=Spring 2=Summer 3=Fall 4=Winter
3 Hour Hour (0 to 23)
4 Holiday Is this Day is holiday or not
5 Day of the Week Day of the week (0 to 6)
6 Working Day Does the day is working day or not
7 Weather Type Weather Type -> 1=Sunny 2=Cloudy 3=Windy 4=Rainy
8 Temperature F Normalized temperature in fehrenhite
9 Temperature Feels F Normalized feeling temperature in fehrenhite
10 Humidity Normalized humidity
11 Wind Speed Normalized wind speed
12 Casual Users Count of casual users
13 Registered Users Count of registered users
14 Total Users Count of total rental bikes including both casual and registered

Tools:

  • Jupyter notebook
  • Libraries(Pandas,Numpy,matplotlib,seaborn,plotly)
  • Git Bash
  • Zoom

Questions:

  • What are the most crowded days and what time was it?
  • What is the difference between registered and casual users?
  • How does the season affect the number of users?
  • Did the low temperature in winter affect the registered users and casual users?

Algorithms:

  • Time Series Data

MVP:

The MVP goal is to answer at least three of the mentioned questions.

Authors