Work Session on Statistical Data Editing International collaboration and processing tools Paris France 2830 April 2014 Saara Oinonen Statistics Finland Contents Background Editing projects and guidelines for editing ID: 360159
Download Presentation The PPT/PDF document "SAS Enterprise Guide project for editing..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
SAS Enterprise Guide project for editing and imputation
Work Session on Statistical Data Editing:International collaboration and processing tools
Paris, France, 28-30 April 2014
Saara Oinonen, Statistics FinlandSlide2
Contents
Background: Editing projects and guidelines for editing
Development of SAS Enterprise Guide project for editing and imputation
Outline of SAS Enterprise Guide project for editing and imputation
Future development
2
UNECE April 2014 France, Saara OinonenSlide3
Background
StatFi has focused on developing editing methods for past five years and in futureProject 1 surveyed editing practices at
StatFi
and researched international development (2009-2011)
→ Process model for editing and imputationProject 2 implements process model, selective editing methods and
editing tools for four pilot statistics (2012-2014)→ SAS EG project for editing and imputation
Project 3 will implement SAS EG guide for editing and imputation for all statistics suitable for selective editing methods (2014-2015)
Two software were chosen for implementing:
SELEKT from Statistics Sweden for selective editingBANFF from Statistics Canada for e.g. error localization and imputation
3
UNECE April 2014 France, Saara OinonenSlide4
Development of SAS EG project for editing and imputation
Basis for implementing selective editing methods was the process model for editing and imputationAll editing features in one programPiloting statistics had suggestions for editing tool
Easy to use, no programming
Priorisaton of units subjected for manual editing
Flexibility on parameter selection etc.
SELEKT and BANFF were incorporated in one SAS EG project
Additional macro programs and parametrisation provide continuous generic program for editing and imputation…
4
UNECE April 2014 France, Saara OinonenSlide5
5
UNECE April 2014 France, Saara Oinonen
Linked programs
Information notes
Datasets created by a program
Run whole project
Run one Process flow
Run one program
Run parts of one programSlide6
SAS EG project for editing and imputation
6
UNECE April 2014 France, Saara Oinonen
Process Flow
includes
info notes and
macro programs
UIR
= User input reguired
→ parameter definitions
Macro programs:
DEF: parameter definitions
PRE: data preparations
MAC: macros for e.g. indicator calculation
BAN: BANFF
SEL: SELEKTSlide7
SAS EG Project for editing and imputation
Parsed programs and codes for different purposes → editing process becomes easier to understandDifferent types of macro programs:
Defining (DEF): parameter definitions, UIR
Preparing (PRE): Data alterations, no UIR
Executive (MAC, BAN, SEL): Produces results or priorisation of units, no UIR
Linked programs and hierarchical levels
Executive programs can be run independently and they can be omitted → suitable for all statistics which need editingBlocks or modules can be separated for lighter operations
7
UNECE April 2014 France, Saara OinonenSlide8
Results from SAS EG project
8
UNECE April 2014 France, Saara Oinonen
SELEKTSlide9
Development proceeds
Imputations methods, further parametrisationManuals, study material, demoConcept library and methodology bank: Database of terms and method wiki available for everyone
User interface
9
UNECE April 2014 France, Saara Oinonen
Thank you for your attention!
Contact: saara.oinonen@stat.fi