FluTE version history

Version 1.0 - January 31, 2009

Used for early versions of Los Angeles County pandemic planning. 
The parallel version runs, but long-distance travel is not implemented 
so the nation-wide simulations should not be run.  School closure is 
permanent – schools don’t re-open after they are closed.

*********************************************************

Version 1.1 - March 13, 2009

Added the childvaccinationpolicy parameter, which can be used to 
vaccinate all children, children enrolled in elementary and middle 
school, or children enrolled in elementary, middle, and high school. 
This was used for the Alachua County simulations.

The option "closeschools" has been changed so that it represents the 
number of days schools will stay closed. For example, setting it to 30 
would mean that they are closed for one month.  

The parallel code to read the tract data has been rewritten so that 
tracts in the same county will be on the same node. This makes it 
easier to associate nodes with airports and to implement county-wide 
policies.

Added short-term long-distance travel to the model based on the 
description in Germann et al. [2006]. 

There is now an option (seedairports) to use the top 15 airports in the 
US as sources of infection.

Added more information to the summary file.

*********************************************************

Version 1.2 - April 2, 2009

Fixed bugs in the prime boost strategies.  These strategies had been
giving boosts to too many people in earlier versions.

*********************************************************

Version 1.3 - April 23, 2009

Moved parameter file reading code to EpiModelParameters class.
Added the antiviralpolicy parameter to the configuration file.

*********************************************************

Version 1.4 - May 12, 2009

Added new antiviral distribution parameters, including treatment only
and full TAP (full TAP not finished yet).
Added contact ascertainment fraction (for full TAP).
Added runlength parameter for setting the number of simulation days.
Added schoolclosurepolicy: set to none for no school closure, all for 
county-wide school closure, and bytractandage for tract closure upon ascertainment.

Ring vaccination might be buggy.

*********************************************************

Version 1.5 - May 21, 2009

Major bug fix - epimodel.cpp was not using the VE and AVE parameters
from the configuration file since version 1.3.  

The Tracts files are now output by mpiflute. 
 
Tract files are only output when Log files are generated.

I took out the dummy data for day 0 in the viral load data, so the 
average case serial interval is 3.6 days instead of 4.6 days.  I had 
to re-calibrate the beta to R0 conversion.  

*********************************************************

Version 1.6 - June 10, 2009

Fixed reactive ring vaccination.

Moved some definitions from params.cpp to epimodelparams.cpp.

*********************************************************

Version 1.7 - July 16, 2009

Fixed R0flute compilation in the Makefile.

Major revision of vaccines.  Data for each vaccine type is stored 
in vaccinedatastruct's.  Added new categories of individuals so that
vaccines can be restricted for these classes of individuals (e.g.,
asthmatics, pregnant people). 

The configuration parameters for vaccines are different now.

Daytime household contacts have been added.  Individuals who are
unemployed and not enrolled in school may interact with other 
household members in the same situation.  Contact probabilities and 
R0 were recalibrated because of this.

*********************************************************

Version 1.8 - August 5, 2009

Fixed bug in mass vaccination when more than one dose is required.
Fixed bug in reactive vaccination strategies for mpiflute.

Fixed Summary and Individuals file output bugs for mpiflute.

Added vaccine delivery schedule parameter for phased vaccination
strategies.

*********************************************************

Version 1.9 - September 1, 2009

Changed vaccine prioritization implementation again.
Changed pregnant, at-risk, and high priority categories.
Added "vaccineefficacybyage".
Added age category of "under 6 months".
Prime-boost strategies might be buggy in this version.

*********************************************************

Version 1.10 - September 21, 2009

Updated README file.
Slight change in the default vaccine buildup assumption.
Changed default values of the vaccination priority groups
so that everyone has the same priority.

*********************************************************

Version 1.11 - November 2, 2009

Added an option for requiring 2 doses for certain age groups.
Added option for pre-existing immunity.

*********************************************************

Version 1.12 - December 21, 2009

Added option to open schools in different states at different times.

Allows one to change vaccine prioritization in the middle of a simulation.

*********************************************************

Version 1.13 - January 14, 2010

Minor optimizations in epimodel.cpp for speed.
Fixed bug in dayinfectsusceptibles that could cause preschoolers in 
playgroups to transmit to family members.

*********************************************************

Version 1.14 - March 22, 2010

Fixed some error and warning messages in mpiflute.

*********************************************************

Version 1.15 - April 26, 2010

Added parameter for the degree of protection for those with
pre-existing immunity (had previously assumed 100%).

Changed random number generator to dSFMT to be more efficient for 
generating doubles.

Used Mark Robien's suggestion to base household composition on PUMS
data instead of the legacy code from EpiCast.  Had to recalibrate
because of the new household size distribution.
