1. Original Visualization
The orignal visualization reflects the death cases and positive cases situation in Jakarta. The data source is from Open Data Covid-19 Provinsi DKI Jakarta.For creating the data visualisation above, the file dated on 30 Juni 2021 Pukul 10.00 is used. The purpose of visualization is to prepare two data visualization to support the news write-up of the latest development in DKI Jakarta.
2. Critiques & Suggestions
2.1 Clarity
Improper annotation for top 5 districts: The annotations of top 5 districts are the district names which can’t reflect the number of deaths and the annotations overlap the point of CENGKENG. Therefore, the annotation didn’t help highlight top 5 districts and even make the visualization effect worse. The information readers can get from annotation is the name of top 5 highest districts which can be found easily on the Y-axis.
Improper title: The title reflect that this chart reflects the cumulative death in top 5 districts. However, the chart contains all the districts and only highlight the top 5 districts by labeling the district names. This choice of words for title is misleading and should not contains top 5 if chart actually reflect all the districts epidemic situation.
- Improper X-axis title: The x-axis title of second visualization is misleading. It will make readers interpret wrongly about the level of cumulative positive cases data. It should also mentioned the data was grouped by city and should also show the variation of average cumulative data.
2.2 Aethetics
Improper Chart Type of Cumulative Deaths: Since the data is cumulative deaths, scatter plot can’t help readers understand the deaths and hard to figure out the accurate number of deaths of different districts. Moreover, the scatter plot is also difficult for readers to find the point corresponding to each district. The scatter plot can’t help readers to find insightful ideas about death cases.
Improper color choice: The chart used blue for scatter plot and bar chart to reflect the cumulative death and epidemic situation. The selection of blue is not appropriate for this dangerous and severe situation. Blue represent the cold and clam which is not as good as red for this situation.
Length of titles: The title for two charts are two lines which are too long for readers to capture the key information for charts. Moreover, the readers will be attracted by the chart and ignore the long title when they first look at the charts. Therefore, the title should be simple and easy to read. We can put the additional information put to the sub-titles.
3. Proposed Design
3.1 Sketch
3.2 Final Design
The address for this visualization on tableau public is here.
3.3 Advantage of Proposed Design
Visualization of map for cities and districts can help readers understand the epidemic situation better in the geographic level. By doing this, the readers can also analyze the epidemic situation by looking at the neighbor sub-districts.
Death rate is also important for covid 19 epidemic situation evaluation. Therefore, the proposed design include the death rate for each district and also label the top 5 districts with highest death rate. This can make the visualization of epidemic situation more comprehensive.
95% upper and lower bar can help readers to understand the variation of cumulative death cases in different cities. In the tooltip, the scatter plot can also help readers understand the distribution of sub-districts for death rate and positive cases.
4. Main Observations
JAKARTA PUSAT: Jakarta Pusat has the lowest average cumulative deaths and variance of sub-districts. The average cumulative positive cases is around 1229. The 95% upper bar is 1390 and the 95% lower bar is 1068. Moreover, the average death rate is 1.8% which is the highest among five cities. This indicate that the medical treatment in Jakarta Pusat.
JAKARTA UTARA: Jakarta Utara has the highest average cumulative deaths and the bar length is the highest amoung five cities. The average cumulative positive cases is around 2117. The average death rate per sub-district is 1.5%. Although the average cumulative positive cases is highest, the district with highest cumulative deaths is in Jakarta Barat. This indicate that the medical treatments in Jakarta Utara are effective for prevent deaths.
JAKARTA TIMUR: From the top 5 districts with highest cumulative deaths, there are two of them are from Jakarta Timur. It has the second highest average cumulative positive cases and second highest average death rate which is 1.7%. Moreover, the 95% bar length is low. Therefore, this indicate Jakarta Timur has generally high number of positive cases and death rates. Jakarta Timur should have more attention from government to control the epidemic situation.
5. Data Visualization Process
5.1 Data Preparation
- The origin dataset is as following:
- Delete the NA values.
- Get the longitude and latitude for districts and cities by Jupyter Notebook.
- Extract clean data for sub-districts.
- Link the two csv data sets together in tableau.
- Change the headers title to English and correct the data type of columns.
5.2 Map - city
- Put the longitude and latitude of city to column and row.
- Change the color and size based on average cumulative positive cases of sub-districts.
- Create a new calculated field.
- Select all in compute using in table calculation.
- Add “rank-city” to Tooltip and add city name to Label. Then we also edit the title.
- Edit the Tooltip to reflect the average cumulative positive cases and rank.
- Map-city is finished.
5.3 Map - district
- Put the longitude and latitude of districts to column and row.
- Drag “City” to color and drag “Died” to size.
- Create a new calculated field.
- Select all for compute using.
- Drag “rank-district” to Label and change the mark label of top 5 districts to “Always Show”.
- Drag the “death rate” and “rank-district” to Tooltip and edit the Tooltip.
- Map-districts is finished.
5.4 Average number of positive cases in cities
- Drag “Positive” to column and change the measure to “Average”.
- Drag “City” to row and order the bar chart based on average positive cases.
- Create the new calculated fields to calculate the 95% upper and lower limit for positive cases.
- Drag the “positive lower bar” and “positive upper bar” to Detail.
- Right click the X-axis and add the reference line.
- Create a new calculated field for horizontal lines between vertical lines.
- Drag “positive lower bar” to column and drag “positive bar length” to Size.
- Right click the X-axis and choose “Dual Axis”.
- Choose the Size for Gantt Bar.
- Edit the Tooltip.
- Edit the title.
- This chart is finished.
5.5 Scatter plot for death rate and positive cases
- Create a new calculated field named death rate for sub-districts.
- Drag variables to column and row. Then drag “Sub-district” to Detail.
Change the color of scatter to yellow.
Add reference line to Y-axis.
- Right click on the reference line and set the “Shading” to 0%.
- This chart is finished. We can add this to the Tooltip of last chart.
5.6 Number of death and death rate in districts
- Drag the variables to column and row.
- Drag “Died” and “death rate” to the color of their chart.
- Right click the X-axis of two charts and add reference line.
- Order the bar chart descending and label the top 5 districts by “Died” and “death rate”.
- Edit the Tooltip.
- Edit the Title.
- This chart is finished.
5.7 Dashboard
Set the title - “Situation of patients diagnosed as positive for COVID-19, Jakarta of Indonesia”.
Add the data source in the bottom of dashboard: “Data Source: Open Data Covid-19 Provinsi DKI Jakarta(https://riwayat-file-covid-19-dki-jakarta-jakartagis.hub.arcgis.com/)”
Arrange the location of charts properly and finish the dashboard.