home *** CD-ROM | disk | FTP | other *** search
- ## -*-Csh-*-
- # ==========================================================================
- # Csh-Example.csh
- #
- # Distributed as an example of Alpha's Csh mode.
- #
- # Csh mode is available at
- #
- # <http://www.princeton.edu/~cupright/computing/alpha/>
- # ==========================================================================
- ##
-
- #!/usr/princeton/bin/tcsh
- # ==========================================================================
- # Corporate Policies Analysis - Exploring 30 years of family benefits
- #
- # FILE: "cps-master.csh"
- # created: 09/25/99 {10:46:22 pm}
- # last update: 08/22/00 {01:22:07 pm}
- # Description:
- #
- # The master source file for beginning analysis of data extracted from
- # the Current Population Surveys CD-ROMs. This source file should be run
- # from the Unix folder benefits/cps/analysis, which assumes that the data
- # files to be analyzed are contained in the folder benefits/cps/data. To
- # use this file, type
- #
- # % source cps-master.csh
- #
- # where % indicates the Unix prompt.
- #
- # Author: Craig Barton Upright
- # E-mail: <cupright@princeton.edu>
- # mail: Princeton University, Department of Sociology
- # Princeton, New Jersey 08544
- # www: <http://www.princeton.edu/~cupright>
- #
- # Copyright (c) 1999-2000 Craig Barton Upright
- #
- # ==========================================================================
- ##
-
-
- ### Preliminaries
-
- # This analysis requires some sharp spikes in memory requirements.
- # For this reason I monitor the directory size, and keep
- # intermediate data files in a scratch directory.
-
-
- unalias rm
-
-
- # Making Stata "nice", which lowers its priority for CPU time.
- # This keeps our friends at CIT happier.
-
- alias stata nice stata -q
-
-
- # Creating symbolic links to scratch files. Each data file is much too
- # large to keep in a personal Unix account. (In fact, they're so large
- # that I have to compress them even in the scratch folder as soon as I can
- # throughout the analysis.) This source file assumes that the extracted
- # data files are already in a scratch directory named "benefits-cps"
-
- mkdir /var/scratch/benefits-cps
-
- rm -f ../data-temp
- ln -s /var/scratch/benefits-cps/ ../data-temp/
-
- mkdir ../data
-
- # Creating additional directory "output"
-
- mkdir ../output
-
-
- # Removing intermediate data files.
- #
- # "rm" is necessary because Stata will not overwrite unless specifically
- # told to do so, and (except for figures), I generally don't.
-
- rm -f ../data-temp/*-1.1.dta*
- rm -f ../data-temp/*-1.2.dta*
- rm -f ../data-temp/*-1.3.dta*
- rm -f ../data-temp/*-1.4.dta*
- rm -f ../data-temp/*-1.5.dta*
-
-
- ### Analysis, part I: reading in data, transforming variables
- #
- # (1-0.do) "Stata-compressing" the marXX.dta data files.
- # (1-1.do) Selects variables to be used in the analysis.
- # Recodes variable responses from "_state" to create "region".
- # Adjusts weights according to instructions in appendix Q.
- # (This involves eliminating some observations from the 1963
- # through 1975 data sets which were assigned negative weights.
- # The output file will indicate how many observations were
- # dropped.) The march files are then removed to save disk
- # space. (This is done at the very end, at which point we can
- # assume that the procedure worked. This same logic continues
- # in the procedures which follow.)
- # (1-2.do) Transforms variables to determine parental status.
- # (1-3.do) Transforms variables to determine occupation status.
- # (1-4.do) Transforms variables to determine industry.
- # (1-5.do) Creating dummy variables
-
- echo "Analysis, part I: reading in data, transforming variables"
-
- stata do cps-1-0.do > cps-1-0.out
- stata do cps-1-1.do > cps-1-1.out
- stata do cps-1-2.do > cps-1-2.out
- stata do cps-1-3.do > cps-1-3.out
- stata do cps-1-4.do > cps-1-4.out
- stata do cps-1-5.do > cps-1-5.out
-
-
- # cleaning up after part I: removing intermediate data files
-
- rm -f ../data-temp/cps*-1.1*
- rm -f ../data-temp/cps*-1.2*
- rm -f ../data-temp/cps*-1.3*
- rm -f ../data-temp/cps*-1.4*
-
-
- ### Analysis, part II: creating crosstab data files
- #
- # (2-1.do) Extracting, merging crosstab information, national data
- # (2-2.do) Extracting, merging crosstab information, regional data
- # (2-3.do) Creating new variables as specified by Erin, national data
- # (2-4.do) Creating new variables as specified by Erin, regional data
- # (2-5.do) Creating figures with new variables, national data
- # (2-6.do) Creating figures with new variables, regional data
-
-
- echo "Analysis, part II: creating crosstab data files"
-
- stata do cps-2-1.do > cps-2-1.out
- stata do cps-2-2.do > cps-2-2.out
- stata do cps-2-3.do > cps-2-3.out
- stata do cps-2-4.do > cps-2-4.out
- stata do cps-2-5.do > cps-2-5.out
- stata do cps-2-6.do > cps-2-6.out
-
-
- ### Analysis, part III: merging the new variables with Erin's data set
- #
- # The "hrs0100.dta.gz" file is from Erin -- it is the one in which all of
- # the new variables will be merged. It is also rather large, so it has to
- # live in data-temp, the scratch directory.
- #
- # (3-1.do) Obtaining the codebook for Erin's data set
- # (3-2.do) Merging the new nariables into Erin's stripped data set
- # (3-3.do) Merging the new nariables back into Erin's original data set
- # (3-4.do) Adding new data, recoding count variables in original data set
- # (3-5.do) Creating figures with new variables
-
- echo "Analysis, part III: merging the new variables"
-
- stata do cps-3-1.do > cps-3-1.out
- stata do cps-3-2.do > cps-3-2.out
- stata do cps-3-3.do > cps-3-3.out
- stata do cps-3-4.do > cps-3-4.out
- stata do cps-3-5.do > cps-3-5.out
-
-
- # making all cps files world-readable on password protected web site
- #
- # Edit the .password/.ht* files to change access users and passwords.
- # The Unix command "ypmatch username passwd" will give
- # encrypted passwords on this system.
-
- echo "making files world-readable on password protected web site"
- chmod -R 755 ../*
-
-
- ### Cleaning up
-
- mv *.out ../output
-
- rm -f ../figures/NMomAll*
- rm -f ../figures/NWomAll*
- rm -f ../figures/NMomWom*
- rm -f ../figures/NParAll*
-
- gzip ../data/hrs0100.dta
-
- echo "current inst-logics directory size:" `du -ksd ../`
-
-
- # Locating errors: Useful if run in background
-
- echo "locating errors ..."
- grep 'r(' ../output/* | cat
- echo "finished looking for errors"
-
-
- # never comment out this last line!!
-
-
- alias rm rm -i
-
- # .