About the Survival Kit
The Survival Kit is primarily intended to fill a gap in the software available to animal breeders who generally tend to use extremely large data sets and want to estimate random effects. Methods of survival analysis have primarily been developed in the area of clinical biometrics where data sets and number of levels of effects are usually smaller.
Although developed by animal breeders for animal breeders, the programs of the Survival Kit could be interesting for people from other areas encountering similar problems of large models and random effects. To make the Survival Kit user-friendly, commands used in the parameter files mimic the SAS command language.
The models supported by the Survival Kit belong to the group of univariate proportional hazards models with a single response time. Fixed and random covariates (discrete and continuous) as well as strata may be included in the analysis. Both types of covariates may be time-dependent. Features include likelihood ratio tests for the full model and any submodel, prediction of (e.g.) median survival times for any given combination of covariates and various bells and whistles for complex models and large data sets.
With version 6 it is possible to estimate variances of two random effects at the same time using a derivative free algorithm.
From Survival Kit v6.1 onwards it is possible to include correlated random effects. This is useful when the two random effects are assumed to be dependent from each other to some degree. In this case the two random effects should have the same number of levels. The correlation coefficient can be assumed to be known (e.g. from a previous analysis) or be estimated at runtime.
The Survival Kit consists of three programs, Prepare (for data preparation), Cox (for analysis under the Cox model) and Weibull (for analysis under a model with Weibull baseline hazard function). Compiled executables for Windows and Linux are provided, together with the source code. A set of support programs is attached to the Survival Kit:
- digamma - program to transform the value of gamma into trigamma
- simul - database simulation for the Survival Kit
- sortSKit - stand-alone R function for sorting of the recoded data file
- SurvKit4R - R interface package for Survival Kit v6.12 with graphical capabilities
An interface package has been developed in R (R Core Team, 2012) to provide graphical capabilities to plot survival curves. The interface function executes all steps (Prepare and Cox or Weibull according to the user specification) in a consecutive manner, including data sorting if necessary. Manual for the R interface is available via the R help command.
The correlated random effects and the R interface is described in the open access publication Mészáros et al. (2013), which is the suggested form of citation of the Survival Kit.
The programs have been written in Fortran 90 and have been tested on PC (using Lahey's Fortran compiler) and on several UNIX platforms. No system routines are used, except a timing subroutine ("second()" for UNIX platform) which can be replaced or switched off without any consequence.
The size of the program may be varied through changes in parameters affecting the maximum number of records and maximum number of levels of effects to be estimated. These parameters could be set using the appropriate keywords in the parameter text files (see manual). The default values are defined in files called prepinclu.f for Prepare and parinclu.f for Cox/Weibull.
How to get the Survival Kit
The programs, documentation and examples are available as a zipped file Survival_Kit_v6.12 (approx. 5.3 MB). Just click on the filename to follow the download instructions. The compressed files contain the Survival Kit, support programs, manual files and a directory with examples.
The zipped folder contains the source code, Makefiles, interface package to R, compiled Linux/UNIX and compiled Windows executable files.
The Survival Kit can be freely used and distributed, provided its use is being credited Mészáros et al. (2013). Use it at your own risk.
The current version is V6.12, written by V. Ducrocq (vincent.ducrocq(at)jouy.inra.fr), J. Sölkner (johann.soelkner(at)boku.ac.at) and G. Mészáros (gabor.meszaros(at)boku.ac.at), 2010. Please report bugs to any of the three authors.
Ducrocq, V. and J. Sölkner (1994): The Survival Kit - a Fortran package for the analysis of survival data. Proc. 5th World Congress on Genetics Applied to Livestock Production, 22 51-52.
Ducrocq, V. (1997): Survival analysis, a statistical tool for longevity data. Paper presented at the 48th Annual Meeting of the European Association of Animal Production, Vienna, 25-28 Aug 1997.
Ducrocq, V. and J. Sölkner (1998): Implementation of a routing breeding value evaluation of dairy cows using survival analysis techniques. Proc. 6th World Congress on Genetics Applied to Livestock Production, 23, 359-362.
Ducrocq, V. and J. Sölkner (1998): The Survival Kit - a Fortran package for the analysis of survival data. Proc. 6th World Congress on Genetics Applied to Livestock Production, 27, 447-448.
Sölkner, J. and V. Ducrocq (1999): The Survival Kit: a tool for analysis of survival data. Workshop on Genetic Improvement of Functional Traits in Cattle - Longevity, May 9-11, 1999, Jouy-en-Josas, France.
Ducrocq, V., Sölkner, J., Mészáros, G. (2010): Survival Kit v6 - A software package for survival analysis. In: 9th World Congress on Genetics to Livestock Production, August 1-6, 2010, Leipzig, Germany.
Mészáros, G., Sölkner, J., Ducrocq, V. (2013): The Survival Kit: Software to analyze survival data including possibly correlated random effects. Computer Methods and Programs in Biomedicine, 110(3): 503-510. DOI:10.1016/j.cmpb.2013.01.010