This github repo is attached to the manuscript Mapping Dynamic Human Sentiments of Heat Exposure with Location-Based Social Media Data. The code can be directly run on the platform CyberGISX at https://cybergisxhub.cigi.illinois.edu/ with confidential. This study aims to explore the possibility to used location-based social media data to map people’s exposure to heat by analyzing nearreal-time social media Twitter.

How to Use this code on CyberGISX?

1. Clone this repo to CyberGISX

2. Data Retrieval

a. If you want to retrieve the full dataset:

i. Generate a ssh key by: ssh-keygen

ii. Retrieve your ssh key and send it to corresponding author

iii. Authorized used will be given the permission to retrive Twitter data on Keeling

b. You can also use the sample data: https://doi.org/10.6084/m9.figshare.21780065

i. Download the dataset and unzip it

ii. Rename the data_sample to data and upload the dataset to CyberGISX

iii. Change corresponding code in the notebook for path and don't run data collection part of the code

3. Use the notebook provided for near real-time analysis of your own/reproduce the result in the manuscript

Notebook description:

1. Chicago_HE_Twitter.ipny -> Used to generate city level heat exposure with real-time twitter data

2. National_HE_Twitter.ipny -> Used to generate national-level heat exposure with real-time twitter data

3. data_collection.ipny -> Use API to retrieve United State Census Bureau ACS data for socioeconomic data

4. analysis.ipny -> post-analysis of national scale heat exposure map

Folder description:

1. Census_track -> Chicago Census Track level shapefile

2. data_explore -> Notebook for prelimery exploration of twiiter data

3. geo/data20000.txt -> heat dictionary

4. geo -> chicago border shapefile

5. output_2021091312 -> the output in the case study in the manuscript, reproduce this with exact same date & time

6. US -> US county shapefile