{smcl}
{* 18May2009}{...}

{title:Steps in creating a dataset}
{p2colset 4 9 9 2}{...}

{p2col:1.}Select observations.{p_end}
{p2col:2.}Select variables.{p_end}
{p2col:3.}Create new variables.{p_end}
{p2col:4.}Add variable labels.{p_end}
{p2col:5.}Add notes to variables.{p_end}
{p2col:6.}Add value labels.{p_end}
{p2col:7.}Rearranging variables.{p_end}
{p2col:8.}Add internal documentation.{p_end}
{p2col:9.}Add a datasignature.{p_end}
{p2col:10.}Save the dataset.{p_end}
{p2col:11.}SUMMARY of key commands.{p_end}

{it:Note}
For naming conventions for datasets, see {helpb wfstyle:Workflow style conventions.}

{title:1. Select observations}

{it:Drop observations based on comparison}

    {cmd:keep if} {it:exp}
    {cmd:drop if} {it:exp}

    For example:

        {cmd: drop if female==1}

{it:Drop observations based on observation number}

    {cmd:keep in} {it:numeric list}
    {cmd:drop in} {it:numeric list}

    For example:

        {cmd: drop in 1/22}

{title:2. Select variables}

    {cmd:keep } {it:variable list}
    {cmd:drop } {it:variable list}

    For example:

        {cmd: drop female year tempvar}
        {cmd: keep var1-var090}

{title:3. Create new variables}

    {cmd:generate }{it:newvar} {cmd:=} {it:exp} [{cmd:if}] [{cmd:in}]
    {cmd:clonevar }{it:newvar} {cmd:=} {it:sourcevar} [{cmd:if}] [{cmd:in}]
    {cmd:replace }{it:newvar} {cmd:=} {it:exp} [{cmd:if}] [{cmd:in}]

{p}Remember that if a variable is new, give it a new name.
Verify that new variables are constructed correctly.
Keep source variables used to create new variables.{p_end}

{title:4. New variables should have a variable label}

    {cmd:label variable} {it:varname} "{it:label}"

    For example:

        {cmd:label var artsqrt "Square root of # of articles"}

{it:Commands for listing variable labels and other information}

    {cmd:codebook} [{it:varlist}] [{cmd:if}] [{cmd:in}] {cmd:, compact}
    {cmd:describe} [{it:varlist}] [{cmd:if}] [{cmd:in}] {cmd:, simple fullnames numbers}
    {cmd:nmlab} {it:varlist}
    {cmd:tab1} {it:varlist}

{title:5. New variables should have notes documenting their provenance}

    {cmd:note} {it:variable name} {cmd:"}{it:variable label}{cmd:"}

    For example:

        {cmd:local tag "pub# truncated at 20 \ wf5-varnotes.do jsl 2008-04-09."}
        {cmd:note pub1trunc: `tag'}

{title:6. Add value labels to all categorical variables}

{it:Step 1: Defining labels} (ideally, 10 characters or shorter)

    {cmd:label define Lyesno 1 1_yes 0 0_no}

{it:Step 2: Assigning labels}

    {cmd:label value wc Lyesno}

{title:7. Rearranging variables.}

    {cmd:aorder} [{it:varlist}]
    {cmd:order} {it:varlist}
    {cmd:move} {it:variable-to-move target-variable}

    For example:

        {cmd:aorder}
        {cmd:order id}

{title:8. Add internal documentation to a dataset.}

    {cmd:label data "Workflow data from Russian ISSP 2002 \ 2008-04-02"}
    {cmd:note: wf-isspru02.dta \ workflow ch 6 \ wf6-save.do jsl 2008-04-05}

{title:9. Add a datasignature.}

    {cmd:datasignature set}

To confirm:

    {cmd:use wf-datasig02, clear}
    {cmd:datasignature confirm}

To change the signature after modifications to the dataset:

    {cmd:datasignature set, reset}

{title:10. Save the dataset}

    {cmd:quietly compress}
    {cmd:saveold mydata, replace}

{title:12. Summary of key commands}

{cmd:keep if} {it:exp}
{cmd:drop if} {it:exp}

{cmd:keep in} {it:numeric list}
{cmd:drop in} {it:numeric list}

{cmd:keep } {it:variable list}
{cmd:drop } {it:variable list}

{cmd:generate }{it:newvar} {cmd:=} {it:exp} [{cmd:if}] [{cmd:in}]
{cmd:clonevar }{it:newvar} {cmd:=} {it:sourcevar} [{cmd:if}] [{cmd:in}]
{cmd:replace }{it:newvar} {cmd:=} {it:exp} [{cmd:if}] [{cmd:in}]

{cmd:label variable} {it:varname} "{it:label}"
{cmd:note} {it:variable name} {cmd:"}{it:variable label}{cmd:"}

        {cmd:local tag "pub# truncated at 20 \ wf5-varnotes.do jsl 2008-04-09."}
        {cmd:note pub1trunc: `tag'}

{cmd:aorder} [{it:varlist}]
{cmd:order} {it:varlist}

{cmd:label data "Workflow data from Russian ISSP 2002 \ 2008-04-02"}
{cmd:note: wf-isspru02.dta \ workflow ch 6 \ wf6-save.do jsl 2008-04-05}

{cmd:datasignature set}
{cmd:datasignature set, reset}

{cmd:quietly compress}
{cmd:saveold mydata, replace}