Skip to main content

Archived Comments for: An object simulation model for modeling hypothetical disease epidemics – EpiFlex

Back to article

  1. Updated Epiflex software available.

    Brian Hanley, UC Davis Microbiology Graduate Group

    21 January 2008

    A new version of the Epiflex software is available by writing to the author at one of these email addresses:

    bphanley@ucdavis.edu ; brian.hanley@ieee.org ; brian.p.hanley@att.net

    Information for version 7.1

    Fixes:

    1. A bug was found that cut contact rates in cells with very large populations per cell. This has been corrected. This had no impact on the models used for the publication.

    User warning:

    When importing into Excel spreadsheets, data will be truncated at roughly 64,000 lines. You will need to break your files up into smaller ones if you execute large models and you use Excel to import files for processing.

    Features:

    1. New report file improvements

    a. .ICX – Infection contagion detail statistics by area. This file is written at the close of processing.

    Record contents: Area name, Host ID, Demographic group, number of other hosts infected by this host.

    b. IDX – Infection demographic statistics by location. This file is optional, configured on the Run Configuration panel. If it is selected it will slow down processing significantly. (Tests show approximately 3-5 times increase in execution time.) It is written at the end of each cycle.

    Record contents: Cycle, Area Name, Location name, cell#, Location population total, Demographic name, #Demographic , #new infections, #incubating, #prodrome, #manifestation, #chronic, #immune

    Notes: List is not in time order within an area or location, only between cycles. Counts after demographic name are counts for the demographic within the location only. Only locations with active infections are shown.

    c. INX - Infectious contacts details. This file is optional, configured on the Run Configuration panel. If it is selected it will slow down processing significantly. (Tests show approximately 3-5 times increase in execution time.)

    Record contents: Cycle, Area Name, Location name, cell #, Infectee id, Infectee demographic group, Disease contracted, Causative contact, Infector id, Infector demographic group, Infector disease stage

    Notes: List is not in time order within an area or location, only between cycles. (non-cellular location cell# = -1) Only new infection events are shown.

    General Note: All files have format contents in a header to the file.

    2. Optional automatic allocation of cells per location.

    If cell count is set to zero in a model for a location, when the model is run it will allocate cells per location based on the maximum number of hosts that can move through a location based on demographic movement cycles and the number of hosts that are specified within each cell.

    Notes: If a complex demographic pattern is specified some degree of over-allocation may occur since timing of demographic movement is not taken into account. This option should not be used for locations that exclusively use the random draw method for population. It is necessary to reset cell counts to zero each time it is desired to auto-allocate.

    3. Return to home cells.

    In the previous version, hosts did not necessarily return to the same cell they did before when they come around to it in their movement cycle again. This had mild skewing effects toward increasing R values in the model because, for instance, an individual would not always return to the same home they did before. This was done to save memory. Now, up to 100 locations can be specified in a movement cycle, and the person will return to the same one each time. As a result of this feature memory requirements have gone up for each host in the system, so optimal performance requires maximum amounts of RAM. This was implemented as a fixed array of 100 long integers because a linked list would be slower. But if you have insufficient RAM then your performance will suffer.

    4. Age of demographic.

    It is now possible to specify an age range for each demographic. Age is incremented on a 365 day calendar. Age specification is not required. Ages are randomly allocated within the age range without intentional skew.

    5. Cross-immunity.

    It is now possible to specify that an age range has some level of fractional immunity to a disease. If a demographic has age range specified, then that demographic will have some fractional degree of resistance to the disease. This will allow changing both how many of an age appropriate demographic become ill, and if they become ill, how severe the illness becomes. The intention of this feature is to allow modeling of age-related previous exposure to the same strain or a similar strain.

    Note: Cross-immunity is how Epiflex now models asymptomatic disease carriers.

    Competing interests

    No competing interests.

Advertisement