The Bucky model uses two main sources of input: the input graph and CDC-recommended parameters.
The input graph contains data regarding demographics, geographic information, and historical case information. For details, see Graph Information.
The Centers for Disease Control and Prevention (CDC) has published pandemic planning scenarios [fDCP+20] which contain recommended parameters describing biological and epidemiological parameters. Of these five planning scenarios, Bucky utilizes scenario 5, which contains the CDC’s current best estimates for disease severity and transmission. These parameters are described in detail, based on information available from the CDC, and summarized in the table below. CDC-recommended parameters are controlled by parameter files located in the par directory.
Bucky Variable Name
Value (Interquartile Range)
Mean generation interval
7.5 (5.5, 8.5)
Mean serial interval
6 (5, 7)
Fraction of infections that are asymptomatic
Relative infectiousness of asymptomatic individuals
Percentage of transmission prior to symptom onset
Case fatality ratio
0-49 years: 0.0005
50-64 years: 0.002
65+ years: 0.013
Case hospitalization ratio
0-49 years: 0.017
50-64 years: 0.045
65+ years: 0.074
TIme from symptom onset to hospitalization
0-49 years: 6 days
50-64 years: 6 days
65+ years: 4 days
Duration of hospitalization
0-49 years: 4.9 days
50-64 years: 7.6 days
65+ years: 8.1 days
Time between death and reporting
0-49 years: 7.1 days
50-64 years: 7.2 days
65+ years: 6.6
The following parameters describe the transmissibility of the virus. The percentage of infections that are asymptomatic, asym_frac, refers to the percentage of infections that will never develop symptoms. This is a difficult parameter to estimate due to logistical complications (individuals would need to be tested to ensure they remain asymptomatic while infectious) and because the level of asymptomatic infections varies by age. The best estimate for this parameter is the midpoint between the lower bound of [BCB+20], the upper bound of [PTC+20], which corresponds to the estimates from [OT20].
The relative infectiousness of asymptomatic individuals compared to symptomatic individuals rel_inf_asym is calculated using upper and lower bounds on the difference in viral dynamics between asymptomatic and symptomatic cases. The lower bound is derived from data indicating that more severe cases have higher viral loads [LYW+20] and a study that indicates symptomatic cases shed for longer and have higher viral loads than asymptomatic cases [NYS+20]. Other studies indicate that both symptomatic and asymptomatic cases have similar duration and viral shedding [LKL+20], which is used as the upper bound.
The final parameter relating to disease transmission is the fraction of transmission prior to symptom onset frac_trans_before_sym which corresponds to the percentage of new cases that were caused by transmission from an individual before they become symptomatic. The lower bound is derived from [HLW+20], with the upper bound derived from [CGM+20].
The mean serial interval, Ts, is the time in days from exposure to onset of symptoms and is taken from [MCH+20]. The mean generation interval, Tg, is the period of time (in days) between symptom onset for one individual and symptom onset for a person they have infected. This value is from [HLW+20].
The case fatality ratio (CFR) is the number of individuals who will die of the disease; the case hospitalization-severity ratio (CHR) corresponds to the number of cases that are severe and necessitate hospitalization. Within the context of the United States, this ratio corresponds to the individuals admitted to a hospital. In a context where access to medical care is limited, this ratio corresponds to the ratio of individuals who exhibit severe disease symptoms.
Hospital-related parameters are derived using data from COVID-Net [CDC] and the CDC’s Data Collation and Integration for Public Health Event Response (DCIPHER). All data is taken from the period between March 1, 2020 to July 15, 2020 unless otherwise noted. The time it takes from symptom onset to hospitalization in days is denoted by I_to_H_time. The number of days an individual will be hospitalized is H_TIME. Finally, the number of days between death and reporting is D_REPORT_TIME.
The Bucky model generates one file per Monte Carlo run. This data is post-processed to combine data across all dates and simulations. It can then be aggregated at desired geographic levels. A separate file is created for each requested administrative level, with each row indexed by data, admin ID, and quantile. The columns of this output file are described in the tables below.
Aggregated files are placed in subfolder named using the Monte Carlo ID within the specified output directory. Filenames are constructed by appending the aggregation level with the aggregation type (quantiles vs mean). For example, the following file contains quantiles at the national level:
An example output directory structure is shown below:
│ └── ADM1
│ ├── adm1_AlabamaDailyReportedCases2020-07-26.png
│ ├── adm1_AlabamaDailyReportedCases2020-08-02.png
│ ├── ...
│ ├── DailyReportedCases_Alabama.png
│ ├── ...
The adm ID corresponding to the geographic level (i.e. adm2 ID)
Case reporting rate
Current number of actively infectious but asymptomatic cases
Cumulative number of cumulative cases (including unreported)
Cumulative number of deaths
Cumulative number of deaths per 100,000 people
Cumulative number of reported cases
Number of reported cumulative cases per 100,000 people
Current number of hospitalizations
Number of current hospitalizations per 100,000 people
ICU bed usage
Current ventilator usage
Number of daily new cases (including unreported)
Number of daily new deaths
Number of daily new hospitalizations
Number of reported daily new cases
Local doubling time as estimated from the historical data
Local effective reproductive number