⭐ 欢迎来到虫虫下载站! | 📦 资源下载 📁 资源专辑 ℹ️ 关于我们
⭐ 虫虫下载站

📄 survan.doc

📁 完整的生存分析Fortran语言源代码。有5个大的功能模块。
💻 DOC
📖 第 1 页 / 共 4 页
字号:
                                  SURVAN                          Version 2.0 August 1990                      Non-Parametric Survival Analysis                  Kaplan-Meier / Log Rank / Gehan-Breslow                 Proportional Hazards / Logistic Regression                              Barry W. Brown                              James Lovato                           M. Elizabeth Rozell                              David Tuttle        This work was supported by grant CA-16672 from the National Cancer        Institute.        Copyright 1988 for             The University of Texas, M.D. Anderson Cancer Center             Department of Biomathematics, Box 237             1515 Holcombe Blvd.             Houston, TX 77030        Contact: BWB at above address or bwb@mdaali.cancer.utexas.edu        SURVAN can be freely copied and (noncommercially) distributed.                            Thanks to Our Supporters     The creation  of this code was  supported in part  by  the  Core Grant     CA11672 of  the  National Cancer  Institute  of NIH to M.  D. Anderson     Hospital.  Other  support was provided by the  State of Texas and  the     Larry  and Pat McNeil  Professorship in Cancer  Research held  by  Dr.     Brown.                                    History         SURVAN was written  in FLECS  (Fortran  Language Extended  Control     Structures)  which is a pre-processor  similar  in  concept to RATFOR.     The original  version of  FLECS   was written by   Terry Beyer  of the     University   of  Oregon and  placed in  the   public domain.  We added     support for Fortran  77  and    some new control  structures.    Those     interested   in acquiring FLECS  should  contact   Barry Brown at  the     address below.         The  Fortran  produced by FLECS  is  as ugly  as  that produced by     RATFOR.  The  code was polished using pol  from  Toolpack1.2, a public     domain package that can be acquired from  NAG or  via anonymous ftp to     lambda.lanl.gov (128.165.4.4).                          Legalities of Use of SURVAN     This code is copyright 1990 for:                The University of Texas M. D. Anderson Cancer Center                1515 Holcombe Boulevard                Houston, TX 77096     Correspondence concerning the code should be addressed to:                Barry W. Brown                Chief, Section of Computer Science                Department of Biomathematics, Box 237     at the above address.         We  want to make  this  program as widely  available  as possible.     Hence, executable  code derived  from the  source provided, the source     itself,  and the accompanying documentation  may  be freely copied and     distributed  provided  that this   copyright section accompanies  such     copies.  We encourage the distribution of this code on bulletin boards     and as shareware.         Modifications and improvements to the code can similarly be freely     copied and distributed.  Barry Brown  would appreciate  a copy of  any     such improvements for incorporation into later versions of SURVAN.         Source  code   may be extracted   and used  for any non-commercial     purpose.  Non-commercial means that at most  a modest distribution fee     is charged, as shareware.  We would appreciate acknowledgment  of such     use of our code.         Commercial use of the source code requires written permission from     the Department of Biomathematics at the above address.  Our  policy is     generally to grant such   permission  provided appropriate  credit  is     given to us in the documentation.                                  NO WARRANTY         WE PROVIDE ABSOLUTELY NO WARRANTY  OF ANY KIND EITHER EXPRESSED OR     IMPLIED,  INCLUDING BUT NOT   LIMITED  TO, THE IMPLIED  WARRANTIES  OF     MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.  THE ENTIRE RISK     AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS  WITH YOU.  SHOULD     THIS PROGRAM  PROVE DEFECTIVE, YOU  ASSUME THE  COST  OF ALL NECESSARY     SERVICING, REPAIR OR CORRECTION.         IN NO EVENT SHALL THE UNIVERSITY OF TEXAS OR ANY  OF ITS COMPONENT     INSTITUTIONS INCLUDING  M. D.  ANDERSON HOSPITAL BE  LIABLE TO YOU FOR     DAMAGES, INCLUDING  ANY LOST PROFITS,  LOST MONIES, OR  OTHER SPECIAL,     INCIDENTAL  OR CONSEQUENTIAL  DAMAGES  ARISING OUT   OF  THE  USE   OR     INABILITY TO USE (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA OR     ITS ANALYSIS  BEING RENDERED  INACCURATE  OR LOSSES SUSTAINED BY THIRD     PARTIES) THE PROGRAM.     (Above NO WARRANTY modified from the GNU NO WARRANTY statement.)                             INTRODUCTION    SURVAN performs the following calculations:(1) Kaplan-Meier estimation of survival functions with confidence    limits on estimates and quantiles.(2) Log rank test of the equality of survival functions of several    groups (optionally over several strata).(3) Gehan  Breslow test  of  the  equality  of  survival functions  ofseveral    groups (optionally over several strata).(4) Proportional hazards model (Cox model) estimation of coefficients    and underlying hazard and survival function.(5) Logistic regression estimation.    In  addition,  SURVAN performs   a few general purpose descriptivetasks that may be useful in setting up the above:(6) Frequency distributions.(7) Condescriptive statistics (mean, sd, min, max, etc.).                              References    This document describes the use of the SURVAN computer program anddoes  not explain  any    of the  statistical   theory behind SURVAN'scalculations.  For  such explanations, the  reader is  referred to thefollowing four references.  The first two are written at an elementarylevel, the following two are advanced.Gross,    A.  J.   and   Clark,  V.A.   (1975) Survival Distributions:Reliability Applications in the Biomedical Sciences.  New York: Wiley.Lee,  E.  T.  (1980)  Statistical Methods  for Survival Data Analysis.Belmont, CA: Lifetime Learning Publications (A Division  of Wadsworth,Inc.)Cox, D. R.  and Oakes, D.   (1984) Analysis  of Survival Data. London:Chapman and Hall.Kalbfleish, J. D. and Prentice, R. L.  (1980) The Statistical Analysisof Failure Time Data.  New York: Wiley.                               Methods    Most of  the methods used  by  SURVAN  are totally standard; thoseused where alternatives exist are noted here.  The handling of ties inthe proportional hazards model is as described in:Breslow, N. E. (1975)  Covariance Analysis of Censored  Survival Data.Biometrics, 30, 89-100.    The underlying hazard and  survival functions in  the proportionalhazards model as estimated by the methods of:Kalbfleish, J. D.  and   Prentice, R. L. (1973)   Marginal likelihoodsbased on Cox's regression and life model.  Biometrika, 60, 267-278.    Confindence  intervals for  Kaplan-Meier  survival curve estimatesare by the method of:Simon,  R. and Lee,  Y.  (1982)  Nonparametric  confidence  limits forsurvival probabilities and  the median.  Cancer Treatment Reports, 66,37-42.                  Other People's Code Used in SURVAN    Wherever possible, we use good code written by experts in place ofreinventing  a poorer version  ourselves.  The  following are used  inSURVAN.   All are  available  by electronic  mail  from netlib.    Forinformation on  netlib,  send electronic mail whose  content is  'sendindex' to netlib@research.att.com.    The general function  minimization  routines  written by David Gayare used in proportional hazards and logistic regression.  The routineused,   DMNH,  is actually an  update   of  the  one  described in thereference.Gay,  D. M.   (1983) Algorithm   611.  Subroutines  for  UnconstrainedMinimization Using a Model/Trust-Region Approach.  ACM Transactions onMathematical Software, 9, 503-524.    The public portion of the PORT library was used for dynamic memoryallocation.Fox,  P. A.,   Hall, A. D.,   and  Schryer,   N. L.  (1978)  The  PORTMathematical  Subroutine Library.  ACM  Transactions  on  MathematicalSoftware, 4, 104-126.    The  zero-finder described  below is  used   in several places  inSURVAN.Bus, J. C. P.  and Dekker, T. J. (1975)  Two Efficient Algorithms withGuaranteed  Convergence  for  Finding a Zero     of   a Function.  ACMTransactions on Mathematical Software, 1, 330-345.                      An Annotated Computer Run                               SURVAN                     Version 1.0 December 1990                  Non-Parametric Survival Analysis                           Barry W. Brown                            James Lovato                         M. Elizabeth Rozell                            David Tuttle   This work was supported by grant CA-16672 from the National Cancer   Institute.   Copyright 1988 for              The University of Texas, M.D. Anderson Cancer Center              Department of Biomathematics, Box 237              1515 Holcombe Blvd.              Houston, TX 77030   Contact: BWB at above address or bwb@mdaali.cancer.utexas.edu   SURVAN can be freely copied and (noncommercially) distributed.Press Return / Enter to continue:?Please enter a name of up to 8 characters for this run of SURVAN.sexampleEnter:   (0) for no echo recording   (1) To echo the entire session to a file   (2) To echo the input only to a file   (3) For functions (1) and (2)? 3Echo of session will be written to file: "sexample.ALLECHO        "Echo of input will be written to file:   "sexample.INECHO         "                      ===== Begin Commentary =====      The  user has an option  to produce a   transcript of the SURVANsession.  Option  1 records  everything  seen  on the screen includinguser prompts and user input.  Option 2 records only the user input.    SURVAN names files using  the  user  supplied  name followed  by aperiod  followed  by a  character string  indicating  the type of filebeing written.   In   MS-DOS,  the length  of  the    character stringfollowing the period is limited to three.                       ===== End Commentary =====Enter:   (1) For results to be displayed at the terminal   (2) For results to be written to a file   (3) For both (1) and (2)? 3Results will be written to file:         "sexample.RESULTS    "Do you want page breaks in your results file? (Y/N)? y                      ===== Begin Commentary =====      Generally but not  always, the  results should  be written  to afile whether or not they are displayed  to the terminal.  The file canbe  printed for reviewing  the  results.   Also,  SURVAN has   no plotcapabilities so  the user might extract  Kaplan Meier estimates with atext editor and plot them using a different program.      Selecting page breaks causes the results  fileto  be broken intopages at logical places.  Page breaks are useful when the results fileis to printed.                       ===== End Commentary =====Do you want to restore a state from a previous run? (y/n)? n                      ===== Begin Commentary =====      SURVAN can read a  state file which  has been produced by a SAVEcommand on  a  previous SURVAN run.   The  state   file  contains  thevariable names, definitions  of survival variable,  group, strata, andcompulsory   variables as they were when   the status was saved.  Theywill be restored if the y option is chosen here.                       ===== End Commentary =====***** ENTER DATA *****Do you want to use an existing binary file (y/n)? nEnter name of binary file to receive a copy of the data.Specify file name or * for terminal output:? sexample.binCreated file sexample.bin --   -- P(roceed) or R(etry file specification)?? p                      ===== Begin Commentary =====    When SURVAN reads a  data file, it produces  a binary (machine butnot human readable) file whose name is specified by the user. Althoughthere is an  option to type * for  terminal output  this option shouldnever be  chosen for a  binary file.  Reading a  binary file is fasterthan reading a raw data file, so if one  is available from  a previousSURVAN run, the user may want to use it.

⌨️ 快捷键说明

复制代码 Ctrl + C
搜索代码 Ctrl + F
全屏模式 F11
切换主题 Ctrl + Shift + D
显示快捷键 ?
增大字号 Ctrl + =
减小字号 Ctrl + -