Embedded Files
Michigan had its first two cases of COVID-19 confirmed on March 10, 2020. This dashboard attempts to show a few of the perspectives one can have when analyzing the spread of COVID-19 across the state of Michigan. This dashboard also aims to inform people of the spread of COVID-19 and the disparities that exist between different groups of the population. The dashboard was auto-updating, using data from a combination of websites and APIs which were then converted into Google Sheets to incorporate it into the dashboard in Google Data Studio. The dashboard compares each county in Michigan with their number of infected cases and deaths, from March 1 to June 17, 2020.
[Source Data from March 1 to June 17, 2020]
In Dashboard #1 below, we present overall statistics about the number of infected cases, death cases, the total number of tests conducted, and positive-case percentage in Michigan versus in the U.S. for the time period March 1 to June 17, 2020. We also illustrate that, overall, Michigan has a higher fatality rate than the U.S. but a lower positive test percentage. The lower positive test percentage means that Michigan is farther along the ‘curve’ of the epidemic. In all of the time-series graphs, we see that both Michigan and the U.S. have decreasing amounts of daily cases and deaths. The data in the dashboard can all be downloaded by hovering over the top right of the visualization.
In Dashboard #1 below, we present overall statistics about the number of infected cases, death cases, the total number of tests conducted, and positive-case percentage in Michigan versus in the U.S. for the time period March 1 to June 17, 2020. We also illustrate that, overall, Michigan has a higher fatality rate than the U.S. but a lower positive test percentage. The lower positive test percentage means that Michigan is farther along the ‘curve’ of the epidemic. In all of the time-series graphs, we see that both Michigan and the U.S. have decreasing amounts of daily cases and deaths. The data in the dashboard can all be downloaded by hovering over the top right of the visualization.
We use the NCOV19 API [2] and the New York Times COVID repository [7] to gather current county-specific data in the U.S., the Coronavirus API [3] to get time-series data for Michigan, and the COVID Tracking Project API [4] to get time-series data for the U.S.. For the numbers regarding Michigan's infections compared to the U.S., we use a simple division of the two numbers we display on each side for cases or deaths to showcase the percentages. For the time-series graph for Michigan, the Coronavirus API provides this data for each state. Lastly, the time-series data for the U.S. is provided by the COVID Tracking Project API and that is visualized to put the Michigan time-series graph in the context of the whole country.
We use the NCOV19 API [2] and the New York Times COVID repository [7] to gather current county-specific data in the U.S., the Coronavirus API [3] to get time-series data for Michigan, and the COVID Tracking Project API [4] to get time-series data for the U.S.. For the numbers regarding Michigan's infections compared to the U.S., we use a simple division of the two numbers we display on each side for cases or deaths to showcase the percentages. For the time-series graph for Michigan, the Coronavirus API provides this data for each state. Lastly, the time-series data for the U.S. is provided by the COVID Tracking Project API and that is visualized to put the Michigan time-series graph in the context of the whole country.
In Dashboard #2 below, we present the spread of COVID-19 across Michigan during three different timelines and the cumulative distribution for June 17, 2020. We see that Wayne county and its adjacent counties to the north have nearly all of the infections for the first timeline. For the second time range, we see the virus starting to spread more around the state, including Kent county (4th largest county by population) in the western part of the state. For the last time-specific map, we see the virus has increased its infections in many counties, especially in southeastern Michigan. For the cumulative data, we see that the order of the first six counties by infections matches the order of the counties if we were to order only by population. The data in the dashboard can all be downloaded by hovering over the top right of the visualization.
In Dashboard #2 below, we present the spread of COVID-19 across Michigan during three different timelines and the cumulative distribution for June 17, 2020. We see that Wayne county and its adjacent counties to the north have nearly all of the infections for the first timeline. For the second time range, we see the virus starting to spread more around the state, including Kent county (4th largest county by population) in the western part of the state. For the last time-specific map, we see the virus has increased its infections in many counties, especially in southeastern Michigan. For the cumulative data, we see that the order of the first six counties by infections matches the order of the counties if we were to order only by population. The data in the dashboard can all be downloaded by hovering over the top right of the visualization.
For the maps of Michigan that have county data, we use the data from the New York Times repository on GitHub [7]. We then use the coordinates for each county from the NCOV19 API [2] to plot it on the map. For the specific date ranges, we select the range in Data Studio and subtract the maximum from the minimum to get the infection spread for each range. For the cumulative map, we use the NCOV19 API data as it only loads the most current data so it is more efficient. For each map, one can use the ‘optional metrics’ button in the top right of each map to see data for deaths and fatality rate among MI counties. The table at the bottom also shows the NCOV19 API data that is visualized in the cumulative map.
For the maps of Michigan that have county data, we use the data from the New York Times repository on GitHub [7]. We then use the coordinates for each county from the NCOV19 API [2] to plot it on the map. For the specific date ranges, we select the range in Data Studio and subtract the maximum from the minimum to get the infection spread for each range. For the cumulative map, we use the NCOV19 API data as it only loads the most current data so it is more efficient. For each map, one can use the ‘optional metrics’ button in the top right of each map to see data for deaths and fatality rate among MI counties. The table at the bottom also shows the NCOV19 API data that is visualized in the cumulative map.
In Dashboard #3 below, we present the distribution of the virus in MI among different demographic groups. We present the graph among ethnic groups with the population proportion to provide context for the infection and death percentages. We can see that during March 1 to June 17, the ‘black or African American’ group has a much higher infection rate when compared to their population proportion. For the graph visualization regarding different age groups, we can see that the number of deaths for each age group increases as age increases. This makes sense as we know that the elderly population is more susceptible to the virus. The data in the dashboard can all be downloaded by hovering over the top right of the visualization.
In Dashboard #3 below, we present the distribution of the virus in MI among different demographic groups. We present the graph among ethnic groups with the population proportion to provide context for the infection and death percentages. We can see that during March 1 to June 17, the ‘black or African American’ group has a much higher infection rate when compared to their population proportion. For the graph visualization regarding different age groups, we can see that the number of deaths for each age group increases as age increases. This makes sense as we know that the elderly population is more susceptible to the virus. The data in the dashboard can all be downloaded by hovering over the top right of the visualization.
For Case distribution among Ethnic Groups, we use the ImportHTML function in Google Sheets to scrape The COVID Tracking Project's Racial Data Dashboard [5] to get all of the data (including population proportions) and then graph it. For the Case Distribution among Age Groups, we upload the XLSX file provided by the state of Michigan [1] that has the current distribution of cases and deaths in Michigan for the selected age groups.
For Case distribution among Ethnic Groups, we use the ImportHTML function in Google Sheets to scrape The COVID Tracking Project's Racial Data Dashboard [5] to get all of the data (including population proportions) and then graph it. For the Case Distribution among Age Groups, we upload the XLSX file provided by the state of Michigan [1] that has the current distribution of cases and deaths in Michigan for the selected age groups.
Building the Dashboard:
Building the Dashboard:
This dashboard represents how one can make an informative visualization of data using a combination of HTML tables and JSON APIs. We use a function from GitHub called ImportJSON [6] which takes a JSON file from a URL and imports it to a spreadsheet. This is helpful as one can input the URL for the API query and have data automatically importing and updating. For the numbers showcased in the dashboard, we simply extract those from the APIs. It is often helpful to use the MAX of a column if the data is time-series so we know that the number we display is the most current (cumulative data can only increase in this instance).
This dashboard represents how one can make an informative visualization of data using a combination of HTML tables and JSON APIs. We use a function from GitHub called ImportJSON [6] which takes a JSON file from a URL and imports it to a spreadsheet. This is helpful as one can input the URL for the API query and have data automatically importing and updating. For the numbers showcased in the dashboard, we simply extract those from the APIs. It is often helpful to use the MAX of a column if the data is time-series so we know that the number we display is the most current (cumulative data can only increase in this instance).
References
References
[1] Coronavirus - Michigan Data, 17 June 2020, www.michigan.gov/coronavirus/0,9753,7-406-98163_98173---,00.html.
[1] Coronavirus - Michigan Data, 17 June 2020, www.michigan.gov/coronavirus/0,9753,7-406-98163_98173---,00.html.
[5] Racial Data Dashboard, The COVID Tracking Project, 17 June 2020, https://covidtracking.com/race/dashboard
[5] Racial Data Dashboard, The COVID Tracking Project, 17 June 2020, https://covidtracking.com/race/dashboard
[6] bradjasper (2019) ImportJSON source code (Version 1.6.0) [Source code]. https://github.com/bradjasper/ImportJSON
[6] bradjasper (2019) ImportJSON source code (Version 1.6.0) [Source code]. https://github.com/bradjasper/ImportJSON
[7] Coronavirus (Covid-19) Data in the United States. GitHub; 2020. https://github.com/nytimes/covid-19-data
[7] Coronavirus (Covid-19) Data in the United States. GitHub; 2020. https://github.com/nytimes/covid-19-data
Student Contributor:
Student Contributor:
Theodore Sweeney
Theodore Sweeney
Undergraduate Major in Data Science and Vocal Performance
Undergraduate Major in Data Science and Vocal Performance
University of Michigan at Ann Arbor
University of Michigan at Ann Arbor
Page updated
Report abuse