ISSS608 2017-18 T1 Assign FOO CELONG RAYMOND/IdentifyGroundZero
From my preliminary analysis, I noticed two sets of symptoms which were manifested by population at different time. I will first gather evidence that there is indeed 2 separate outbreaks.
I first find people that posted more than one microblog post and had posted at least two different post containing 2 different symptoms. Then I plot the percentage of each symptom being mentioned with reference to a different symptom. From above chart alone, I can tell that the diarrhea, vomit and nausea symptom is not related to the symptoms in the first outbreak since only a minute percentage of people with symptoms of the first outbreak developed the symptoms of the second outbreak.
From the next chart, I can also guess that flu and pain are likely to be symptoms that developed in the later stage of the first outbreak as they are also weakly related to the diarrhea, vomit and nausea symptoms.
I know that the first 7 symptoms that were first observed on 18-May are:-
- Aches
- Breathing problem
- Chills
- Cough
- Fatigue
- Fever
- Sweat
Downtown is the ground zero of the outbreak because it has the highest number of post across all 7 symptoms. It is also observed that the top 5 city zones are in the same order across all symptoms. Downtown have the highest number of posts followed by uptown, eastside, lakeside and finally suburbia.
The most affected areas seem to form a cone shape which started from downtown and spreads out towards the east and becomes more diffused as it spreads out. All top 5 city zones in the top with the highest number of symptoms are within the boundary of the cone.
Symptoms that were first observed on 19-May are:-
- Diarrhea
- Flu
- Nausea
- Pain
- Vomit
From the chart, I can see that westside and plainville are the two zones with the highest counts of posts across all 5 symptoms except for flu. I observe that for flu, the top 5 zones are the same as those symptoms presented in the 18-May which suggests that flu is a late occurring symptom from the outbreak in the 18-May.
The most affected areas seem to form a cone shape which started from north east of plainville and south east of westside, and spread towards the south west, also becoming more diffuse as it spread south west.
