Midterm_Running_Project

Running data - for midterm

Description

Ultra Marathon running data cleaned for the top 100 distance races in the United States. Data depicts specific athletes, years, events, distance, state of event, and race times as an example of the columns contained in the information. The data shows the top 100 events data from 1963 to 2021.

Project Proposal

• Does Elevation average of states or elevation gain have a factor in race times?
• Do race times get better over the years from 1963 to 2022?
• What are the states with the average race paces that are higher?
• Is age associated to running faster with elevation gain?
• What is the fastest running time per minimum and maximum elevation gain by state?
• Find the correlation between age and race location(state) along with the time it takes to run the races (hours). Is there a correlation?

Table of Contents

• Installation
• Usage
• Contributing
• License
• Contact

Installation

Prerequisites

• List of libraries required:
  • pandas as pd
  • numpy as np
  • matplotlib.pyplot as plt
  • seaborn as sns
  • plotly.express as px
  • import re
  • import folium

Steps

1. The data is very large and can be downloaded from: https://www.kaggle.com/datasets/aiaiaidavid/the-big-dataset-of-ultra-marathon-running The data is pulled from multiple sources including mains websites from:
   • https://ultrasignup.com
   • https://ultrarunning.com
2. Gather further information in regards to the data referenced you want to learn about. I chose the top 100 races by distance (Not by time i.e. a 24 hour run) and only in the United States. I then added elevation gain as well as the highest elevation point of each race. I also included the state each event was located in. For this I just used open source search engines. I have added my newfound data to the data files.
3. Merge the data to clean.

Usage

Descirbe how I cleaned and transformed the data

• To fully understand how the data was cleaned, open the cleaning_data.ipynb in the notebook folder.
• The data from kaggle had over 7.4 million rows of data and only events within the United States were returned.
• New data was added from the top 100 events by participants within the United States. 100 events were chosen as a pool from over 6,000 events from the United States. The new data contains elevation gain, highest elevation of each race, state of race, miles of each race, and average elevation of each state.
• After the data was merged, all events not within the top 100 chosen were removed. All 100 races are by distance and not a timed race.
• The new data set removed nearly 7 million rows and left with roughly 420,000 rows for the top 100 events by distance.
• The data was further manipulated from issues with the dates. Regex was utilized to fix multiple dates formatted in separate ways so a Date Time Group (DTG) could be utilized.

What visiualizations do we need to understand your topic?

• Scatter Plots
• Bar Charts
• Line Plots
• Box Plots
• Choropleth
• Folium Map
• Correlation Matrix

What visualizations display your results?

• Race Locations on Map

• Race Locations on Heat Map

Does Elevation average of states or elevation gain have a factor in race times?

• Average Elevation Gain and Race Pace by State

• Average State Elevation and Race Pace by State

• Average Elevation Gain and Max Pace by State

• Min and Max Pace per Year by Count of Athletes

Do race times get better over the years from 1963 to 2022?

What are the states with the average race paces that are higher?

• Map of Average Race Pace per Events

Is age associated to running faster with elevation gain?

• Scatter Plot of Pace vs Age by Elevation Gain

What is the fastest running time per minimum and maximum elevation gain by state?

• Min and Max Elevation Gain by State

• Fastest Running Times by State for Min and Max Elevation Gain

Find the correlation between age and race location(state) along with the time it takes to run the races (hours). Is there a correlation?

• Age Vs Hours Ran

• Hours Ran Distribution by State

• Average Hours Ran per State

• Correlation Matrix of Pace, Age, and State

• Correlation Matrix of Main Columns

Present key insights

• Elevation Gain and State Elevation has an impact on running times regardless of Age.
• Minimum race pace over time decreases with an increased amount of runners.
• Maximum race pace over time increases with increased amount of runners.
• The highest average race pace in the United States that is highest is in Illinois and Maine.
• Age is associated to running faster with elevation gain, however, with maximum elevation gain age is not a factor.
• The fastest running time per minimum and maximum elevation is in North Carolina (min elevation gain) and Virginia (Max elevation gain).
• There are no direct correlations for age, race location, and hours ran in a race. There are more positive correlations between hours ran, distance, and Elevation.

Tell us what you recommend

• Recommend breaking out the data by distance due to ultra marathon races being anywhere from 26.3 miles and up. i.e. data by 50km races to 100 mile races.

Tell us what future areas you might want/need to study

• Consider studying Terrain type of events, i.e. desert, mountain, forest, beach and seasons to see if that affects runners performance.

Contributing

To contribute to this project, continue to add where to find pertinent data about each race to enhance further exploration of data analysis of Ultra Marathon Running. Continue to provide latitude and longitude of all races for better visuals on maps.

License

This project is licensed under the LGPL-2.1 License.

Contact

• Author - Justin Welch
• GitHub - https://github.com/jdwelch88