America Addresses Two Epidemics Dataset – Cannabis and Coronavirus and their Interactions: Combined Geospatial and Causal Inference Study

Research Hypothesis - That cannabis may predispose to COVID-19 viral infection due to its immunomodulatory, envorinmental contaminants, vaping and smoking inhalation actions. Data was analyzed by geospatial and causal inference techniques in R. Data was gathered from publicly available on line sources including: Data were downloaded from Publicly available datasets including: • US Census bureau 2019 • Five Year American Community Survey 2013-2018 • National Survey of Drug Use and Health (NSDUH) • NSDUH Resticted Use Data Analysis System (RDAS) • US Department of Transport International Flight Data • Worldometer Covid -19 Dataset Data were collected in the six domains of: • COVID numbers • Fights – numbers of flights and numbers of overseas destionations • Median household income • State ethnic composition • Population and population density • Drug use Inverse probability weights were constructed by inverse probability weighting conducted in package ipw in R. Geospatial weights were constructed in package spdep in R.

研究假设——大麻可能会使个体易感染新冠病毒（COVID-19），其关联机制涉及大麻的免疫调节特性、相关环境污染物暴露以及电子烟与吸烟的吸入行为。本研究采用R语言中的地理空间分析与因果推断技术对数据开展分析。
数据均取自公开在线数据源，具体包括：
• 美国人口普查局2019年数据集
• 2013-2018年美国社区调查五年期数据
• 全国药物使用与健康调查（National Survey of Drug Use and Health, NSDUH）
• 全国药物使用与健康调查受限使用数据分析系统（Restricted Use Data Analysis System, RDAS）
• 美国交通部国际航班数据
• Worldometer新冠病毒数据集
数据涵盖以下六大维度：
• 新冠疫情统计数据
• 航空运输数据——航班数量与海外目的地数量
• 家庭收入中位数
• 各州族群构成情况
• 人口总量与人口密度
• 药物使用情况
研究通过R语言的ipw包执行逆概率加权操作以构建逆概率权重，同时借助R语言的spdep包构建地理空间权重。