Transcript Slide 1

STATA
Third group training course in application of information
and communication technology to production and
dissemination of official statistics
10 May – 11July 2007
Gereltuya Altankhuyag, Lecturer/Statistician, UNSIAP
[email protected]
7/21/2015
1
Basic commands of graphics

The commands that draw graphs are:
command
description
--------------------------------------------------------------------graph twoway
scatterplots, line plots, etc.
graph matrix
scatterplot matrices
graph bar
bar charts
graph dot
dot charts
graph box
box-and-whisker plots
graph pie
pie charts
other
more commands to draw statistical graphs
---------------------------------------------------------------------
7/21/2015
2
Basic commands of graphics

The commands that save a previously drawn graph,
redisplay previously saved graphs, and combine
graphs are:
command
description
---------------------------------------------------------------------graph save
save graph to disk
graph use
redisplay graph stored on disk
graph display
redisplay graph stored in memory
graph combine
combine graphs into one
----------------------------------------------------------------------
7/21/2015
3
Basic commands of graphics

The commands for printing a graph are:
command
description
---------------------------------------------------------------------graph print
print currently displayed graph
set printcolor
set how colors are printed
graph export
export .gph file to PostScript, etc.
----------------------------------------------------------------------
 All graphs have extension “graph”
7/21/2015
4
1.5e-04
1.0e-04
0
5.0e-05
Density
 We can generate plots,
e.g. a histogram, with
STATA:
hist pcexp
which shows
per capita expenditure
distn
as skewed
2.0e-04
Basic commands of graphics
0
7/21/2015
20000
40000
Per capita expenditure per year
60000
5
.8
.6
.4
0
.2
We can induce
normality by
transforming
consumption on a log
scale:
gen logpcexp=
ln(pcexp)
hist logpce
Density

1
Basic commands of graphics
7
7/21/2015
8
9
logpcexp
10
11
6
Basic commands of graphics
graph bar (mean) pcexp indsave, over(region) title(Per
capita expenditure and individuals' savings by region)
0
2,000
4,000
6,000
8,000
Per capita expenditure and individuals' savings by region
Dhaka
Chittagon
mean of pcexp
7/21/2015
Khulna
Ragfhahi
mean of indsave
7
10
Household size
To create a box plot of
the household size
vaiable in the hh.dat
dataset:
15
20
Basic commands of graphics
0
5
graph box famsize
7/21/2015
8
Creation of “Do-files” and use of
“Do-file Editor”

Up until now we have used Stata interactively.

We can do batch processing using Do Files.

Do files – a text file containing commands and
instruct Stata to execute the commands stored
in that file.
7/21/2015
9
Creation of “Do-files” and use of
“Do-file Editor”

We can create Do Files using “New Do-file Editor”
7/21/2015
10
Creation of “Do-files” and use of “Do-file Editor”
Do-file Editor has menu bar
Do-file Editor has tool bar
7/21/2015
11
Creation of “Do-files” and use of “Do-file Editor”


The do-file editor lets you submit several commands
to Stata at once
General steps of using the do-file:
 Click on the Do-file Editor to open do-file editor
7/21/2015
12
Creation of “Do-files” and use of
“Do-file Editor”

7/21/2015
Then type the commands that you wish to
submit to Stata
13
Creation of “Do-files” and use of
“Do-file Editor”

7/21/2015
Then click on “Do current file”
14
Creation of “Do-files” and use of “Do-file Editor”
If you wish to execute a subset of the lines in your do-files,
highlight those lines and click “Do current file” button.
7/21/2015
15
Creation of “Do-files” and use of “Do-file
Editor”

Tool bar button “Run current file” is the same
as “Do current file”, but Run suppressess
output. It is unlikely that you will ever need to
use “Run current file” button.
7/21/2015
16
Creation of “Do-files” and use of
“Do-file Editor”


You can put comments in your Do-file
There are ways to include comments in a Dofile:





7/21/2015
Begin with asterisk *
Begin with asterisk and forward slash */
Place the comment in /*…..*/ delimiters
Place the comment after 2 forward slashes //….//
Place the comment after 3 forward slashes ///…///
17
Creation of “Do-files” and use of
“Do-file Editor”
7/21/2015
18
Creation of “Do-files” and use of
“Do-file Editor”
Advantage of using /* */ as a comment indicator
The /* */ comment can be put anywhere – at
the end of a line, even at the middle of a line.
Advantage of using * as a comment indicator
The ‘*’ comment indicator can be used interactively.
The /* */ can be used only in do-files.
7/21/2015
19
Creation of “Do-files” and use of
“Do-file Editor”
 By default Stata assumes that each command
is ended by the carriage return (cr) (ENTER
key press).
 If a Syntax with commands is too long to fit
in one line, we can handle it in 2 ways:
 Using #delimit
 Using comment signs
7/21/2015
20
Creation of “Do-files” and use of “Do-file Editor”
 using #delimit – each commandmust end with
semicolon (;)
7/21/2015
21
Creation of “Do-files” and use of “Do-file Editor”
 using comment signs (*; */; /*..*/; //..//; ///…///)
7/21/2015
22
Creation of “Do-files” and use of “Do-file
Editor”

A do-file stops executing when:
The end of the file is reached
 An exit is executed
 An error occurs


If an error occurs, the remaining commands in
the do-file are not executed
7/21/2015
23
Creation of “Do-files” and use of “Do-file
Editor”

Retrieve saved do-file:




Save do-file:




Click do-file editor
File
Open
Type in your commands in do-file editor
File
Save or Save as
Saved file takes “.do” extension
7/21/2015
24
Creating “log-files”




Stata can record your session into a file called
a “log file”
Log file does not start automatically
You must tell Stata to record your session
By default, log files are created in a format
called Stata Markup and Control Language
(SMCL)
7/21/2015
25
Creating “log-files”

To create log files using do-files:
 Type the followings:
 clear
 capture log close

Specify the location of your log-file in your dofile:
 log using C:\intropov\logfiles\day2, replace
 Specify the start date and time of running do-file:
 di "Job started at $S_TIME on $S_DATE"
7/21/2015
26
Creating “log-files”
Write all commands and comments accordingly
 Specify the finishing date and time of running
do-file:
di "Job ended at $S_TIME on $S_DATE“

 Tell Stata that log is finished:
log close
 Saved file takes “.log” or”smcl” extension
Note: you can make changes in your do-file!!!
7/21/2015
27
Creating “log-files”
 To run do-file which contains log:
 Click on Edit of the do-file menu bar
 Click Select All: all commands and
comments will be highlighted
 Click “Do current file” button on tool bar
7/21/2015
28
Creating “log-files”
7/21/2015
29
Creating “log-files”
7/21/2015
30
Creating “log-files”

To open/retrieve log file click “begin log”
button on toolbar
7/21/2015
31
Creating “log-files”

Find you log-file name in relevant folder
7/21/2015
32
Creating “log-files”

Log files have two extensions:


7/21/2015
filename.smcl – Stata produces this file
when you open log
filename.log – You can load into your word
processor. Do-file uses this extension!!!
33
Creating “log-files”

Choose the extension of your log-file by
scrolling down Save as type:
7/21/2015
34
Creating “log-files”


Then Stata log options window appears and
choose “View existing file (read only)”
The click OK button.
7/21/2015
35
Creating “log-files”

Then you will see all your command records and
outputs.
7/21/2015
36
Creating “log-files”
Meaning
Syntax
To start a log (in smcl)
log using filename
To stop temporarily logging
log off
To resume log
log on
To stop logging and close
the file
log close
To create filename.log
translate filename.smcl
filename.log
7/21/2015
37
Creating “log-files”

Append to an existing log-file:
log using filename, append
 Replace an existing log-file:
log using filename, replace
7/21/2015
38
Use of “ado-files”

Stata has two types of commands in terms of
use:




Commands built into Stata
Ado-file commands
An ado-file is an ASCII text file that contains a
Stata program.
Ado-files usually come with help files.
7/21/2015
39
Use of “ado-files”

To determine your command:
which command name
which ci
 . which ci
 \\Unitednations\Stata9\ado\base\c\ci.ado
 *! version 3.3.15 17mar2005
Which summarize
 . which summarize
 built-in command: summarize
7/21/2015
40
Use of “ado-files”

You can view ado-files:
. viewsource ci.ado
7/21/2015
41
Use of “ado-files”

To know where are ado-files on your PC:
sysdir
sysdir
STATA:
UPDATES:
BASE:
SITE:
PLUS:
PERSONAL:
OLDPLACE:
7/21/2015
\\Unitednations\Stata9\
\\Unitednations\Stata9\ado\updates\
\\Unitednations\Stata9\ado\base\
\\Unitednations\Stata9\ado\site\
c:\ado\plus\
c:\ado\personal\
c:\ado\
42
Use of “ado-files”

To run ado-files:
ado-filename varname
 Examples:
ci pcexp
gini pcexp
lorenz pcexp
7/21/2015
43
Use of “ado-files”

Examples:
. ci pcexp
Variable |
Obs
Mean Std. Err.
[95% Conf. Interval]
-------------+--------------------------------------------------------------pcexp |
519 6901.953 192.7895
6523.207 7280.698
. gini pcexp
Gini coefficient .26204428
7/21/2015
44
Use of “ado-files”
. lorenz pcexp
7/21/2015
45
Inputting data
To input data into Stata, you can use:
7/21/2015
46
Inputting data
edit and input
Enter data from keyboad
insheet
Reads text (ASCII) files created by a
spreadsheet
infile (free format)
Read unformatted ASCII (text) data
infile (fixed format) or Read ASCII (text) data in fixed format with a
infix (fixed format)
dictionary
xmlsave
7/21/2015
Allows datasets to be saved or used in XML
and EXCEL format
47
Inputting data
odbc (open darabase
connectivity)
Read data from ODBC source
sdasave
Read datasets in SAS SPORT format
haver
Read data in Haver Analytic fromat
Stata-Transfer
program
7/21/2015
48
Inputting data

Suppose we have an ASCII text dataset as shown below:
7/21/2015
49
Inputting data
Suppose we have a dataset countries.txt in
c:\intropov\data\ folder
 Use the infile command with syntax
infile variable-list using filename.raw
 In particular, enter
cd c:\intropov\data
infile y_a y_b y_c using countries.txt

7/21/2015
50
Inputting data


When using the infile command for character
(string) variables (eg.hhcode), we need to identify
these variables. For instance
infile y_a y_b y_c str30 domain using countries.txt
For more details regarding infile, enter
help infile1
7/21/2015
51
Inputting data

Suppose we have a dataset try2.txt in
c:\intropov\data folder with the data in
specific fields
Assumes last line is
blank line
7/21/2015
52
Inputting data

Use the infix command to input into Stata
infix domain 1 hhcode 2 age 3-4 using try2.txt,
clear

Infix accepts an “if” statement, e.g.,
infix … if domain ==1
7/21/2015
53
Inputting data
IMPORTING A WORKSHEET
 Suppose we have an Excel worksheet called
countries2.xls also in the c:\intropov\data folder.
 Read the file with Excel, then SAVE AS a comma
delimited CSV file.
 Enter
insheet using countries2.csv, clear
 Note: Stata cannot read/import data directly
from Excel!!
7/21/2015
54
Inputting data
Thus, Stata can read text files with
 Infile (if the data in text is separated by spaces
and does not have strings, or if strings are just
one word, or if all strings are enclosed in
quotes)
 Infix (fixed format text)
 Insheet (if text file was created by a
spreadsheet or db program)
7/21/2015
55
Exporting data

The command outfile save the data in ASCII;
outsheet in a comma delimited format.
File ► Export
►Comma or tabseparated data
7/21/2015
56
Inputting and exporting datasets

There are third party programs, esp.
Stat/Transfer and DBMS/COPY, that perform
translations from one data format (e.g., Excel,
SAS, SPSS, Stata) to another.
7/21/2015
57
Inputting and exporting datasets
7/21/2015
58
Inputting data
There are 2 ways of creating a new dataset in Stata:
1. Use of Edit:
1.1 Open the STATA spreadsheet editor
7/21/2015
59
Inputting data
1.2 Enter data into the editor, when done close
the editor.
1.3 Save dataset
7/21/2015
60
Inputting data
2. Use of input command
2.1 input hhcode pid age
2.2 enter values of variables
2.3 end
2.4 Save dataset
7/21/2015
61
Inputting data
NOTE
 We can copy and paste values from a
spreadsheet
 Data can be entered variable-by- variable or
observation-by-observation
 When entering data variable-by-variable, use
TAB; for observation-by-observation use
ENTER key
 Missing data code is a period “.”
7/21/2015
62
Inputting data
NOTE
 Numerical values SHOULD NOT include
commas; else these are considered string.
 String variable values can be almost any
combination of letters, numbers, symbols up
to 80 characters long in Intercooled/ Small and
up to 244 characters in STATA/SE.
7/21/2015
63
Summary of STATA
session
7/21/2015
64
Summary

Stata Files
.log
.do
records commands and outputs
contains a sequence of Stata
commands for batch execution
.dta
data files
.graph graphics files
.ado for writing new commands
7/21/2015
65
Summary

Basic Commands
Interrupt Stata: Ctrl Break or Stop button
 Exit from Stata: type exit or from File menu,
choose exit or clear
 Help
 Search

7/21/2015
66
Summary

To inspect datasets:









codebook: display contents of variables
count: counts cases
describe: display properties of variables
summarize: descriptive stats
list: show values
tabulate: one or two way freq tables
table: higher dim freq tables
tabstat: tables with statistics
Note: weigh is important!!!!
7/21/2015
67
Summary

To create and change varaibles:
generate/egen: create new variable
 replace: replace old values
 recode: change values by conditions
 label: label var (up to 80 chars)

7/21/2015
68
Summary

To reorganize datasets:
rename: change the name of a variable
 drop: eliminate variables or observations
 keep: keep variables or observations
 clone: copy of an existing variable
 sort: arrange into ascending order
 gsort: arrange in ascending or descending
order

7/21/2015
69
Summary

To reorganize datasets:
append: append two datasets
 merge: merges two datasets
 reshape: converts dataset from wide to long
and vs

7/21/2015
70
Summary

Survey Data Analysis

svy commands
 svyset
 svymean
 svytotal
 svytab
 syyprop
 svyreg
7/21/2015
71
Summary

Importing Datasets
Files must be in text delimited form (.txt)
 Use infile function to import:
infile var1 var2 etc using filename.txt, clear
 Or the infix or insheet functions
 Or use StataTransfer software

7/21/2015
72
END OF STATA SESSION
Please perform EXERCISE 4
7/21/2015
73