Programmability in SPSS 14 SPSS 15 and SPSS

Programmability in SPSS 14, SPSS 15 and SPSS 16 The Revolution Continues Jon Peck Technical Advisor SPSS Copyright (c) SPSS Inc, 2007

Agenda § Recap of SPSS 14 Python programmability § Developer Central § New features in SPSS 15 programmability Writing first-class procedures § Updating the data § New features in SPSS 16 programmability § Interacting with the user § Q&A § Conclusion Copyright (c) SPSS Inc, 2007 §

Quotations from SPSS Users "Because of programmability, SPSS 14 is the most important release since I started using SPSS fifteen years ago. " § "I think I am going to like using Python. " § "Python and SPSS 14 and later are, IMHO, GREAT!" § "By the way, Python is a great addition to SPSS. " § From Info. World (April 19, 2007) § "Of all the tools fueling the dynamic-language trend in the enterprise, general-purpose dynamic languages such as Python and Ruby present the greatest upside for enhancing developer productivity. " Copyright (c) SPSS Inc, 2007 §

The Combination of SPSS and Python SPSS provides a powerful engine for statistical and graphical methods and for data management. § Python® provides a powerful, elegant, and easyto-learn language for controlling and responding to this engine. § Together they provide a comprehensive system for serious applications of analytical methods to data. Copyright (c) SPSS Inc, 2007 §

Programmability Features in SPSS 14, 15, and 16 § SPSS 14. 0 provided § § § SPSS 15 adds Read and write case data Create new variables directly rather than generating syntax Create pivot tables and text blocks via backend API's Easier setup SPSS 16 will add § § § EXTENSION command for user procedures with SPSS syntax Dataset features for complex data management Ability to use R procedures within SPSS through R Plug-In Copyright (c) SPSS Inc, 2007 § § § Programmability Multiple datasets Variable and File Attributes Programmability read-access to case data Ability to control SPSS from a Python program

Programmability Advantages Makes possible easy jobs that respond to datasets, output, environment § Allows greater generality, more automation § Makes jobs more robust § Allows extending the capabilities of SPSS § Enables better organized and more maintainable code § Facilitates staff specialization § Increases productivity § More fun Copyright (c) SPSS Inc, 2007 §

Programmability Overview § Python extends SPSS via § § § Runs in "back-end" syntax context (like macro) § § Sax. Basic scripting runs in "front-end" context Two modes Traditional SPSS syntax window § Drive SPSS from Python (external mode) § § Optional install (licensed with SPSS Base) Copyright (c) SPSS Inc, 2007 § General programming language Access to variable dictionary, case data, and output Access to standard and third-party modules SPSS Developer Central modules Module structure for building libraries of code

Legal Notice § Copyright (c) SPSS Inc, 2007 SPSS is not the owner or licensor of the Python software. Any user of Python must agree to the terms of the Python license agreement located on the Python web site. SPSS is not making any statement about the quality of the Python program. SPSS fully disclaims all liability associated with your use of the Python program.

The SPSS Programmability Software Development Kit § Supports implementing various programming languages § § Requires a programmer to implement a new language VB. NET Plug-In available on Developer Central Works only in external mode Copyright (c) SPSS Inc, 2007 §

How Programmability Works Python interpreter embedded within SPSS § SPSS runs in traditional way until BEGIN PROGRAM command is found § Python collects commands until END PROGRAM command is found; then runs the program § Python can communicate with SPSS through API's (calls to functions) § § Includes running SPSS syntax inside Python program Includes creating macro values for later use in syntax § Python can access SPSS output and data § OMS is a key tool Copyright (c) SPSS Inc, 2007 §

Example: Summarize Categorical Variables BEGIN PROGRAM. import spss, spssaux. Get. SPSSInstall. Dir("SPSSDIR") spssaux. Open. Data. File("SPSSDIR/employee data. sav") DESC !cat. Vars. Run Copyright (c) SPSS Inc, 2007 # find categorical variables cat. Vars = spssaux. Variable. Dict(variable. Level=['nominal', 'ordinal']) if cat. Vars: spss. Submit("FREQ " + " ". join(cat. Vars. variables)) # create a macro listing categorical variables spss. Set. Macro. Value("!cat. Vars", " ". join(cat. Vars. variables)) END PROGRAM.

Programmability Inside or Outside SPSS § Two modes of operation § SPSS Drives mode (inside): traditional syntax context BEGIN PROGRAM …program… END PROGRAM § Program in 14, 15, or 16 is in Python or, new in 16, in R § § X Drives mode (outside): e. Xternal program drives SPSS § Output sent as text to the application – can be suppressed Has performance advantages § Build programs with an IDE § § Even if to be run in traditional mode Copyright (c) SPSS Inc, 2007 Python interpreter (or VB. NET) § No SPSS Viewer, Data Editor, or SPSS user interface §

Python. Win IDE Controlling SPSS (e. Xternal Mode) Copyright (c) SPSS Inc, 2007

Python Resources § Be productive quickly § Get more return as you learn more § Python. org § Python Tutorial § Cheeseshop over 2200 packages as of April 11, 2007 § SPSS Developer Central § SPSS Programming and Data Management, 4 th ed, 2006. Copyright (c) SPSS Inc, 2007 §

Python Books § Dive Into Python book or PDF § Practical Python by Magnus Lie Hetland § Extensive examples and discussion of Python Cookbook, 2 nd ed by Martelli, Ravenscroft, & Ascher § Python in a Nutshell, 2 nd ed by Martelli, O'Reilly § § Very clear, comprehensive reference material wx. Python in Action by Rappin and Dunn § Explains user interface building with wx. Python Copyright (c) SPSS Inc, 2007 §

Cheeseshop: scipy § scipy 0. 5. 2 Scientific Algorithms Library for Python § Scipy. org scipy is an open source library of scientific tools for Python. scipy gathers a variety of high level science and engineering modules together as a single package. scipy provides modules for statistics, optimization, integration, linear algebra, Fourier transforms, signal and image processing, genetic algorithms, ODE solvers, special functions, and more. scipy requires and supplements Num. Py, which provides a multidimensional array object and other basic functionality. § Python is becoming a major language for scientific computing Copyright (c) SPSS Inc, 2007 §

SPSS Developer Central is the web home for developing SPSS applications § Python, . NET, R Integration Plug-Ins § Supplementary modules by SPSS and others § Articles on programmability and graphics § Forums for asking questions and exchanging information § Programmability Extension SDK § Get Python itself from Python. org or CD § § § SPSS 14, 15 use 2. 4. (2. 4. 3) SPSS 16 will use 2. 5 Not limited to programmability § § GPL graphics User-contributed code Key Supplementary Modules spssaux spssdata New for SPSS 15 trans extended. Transforms rake pls enhanced tables. py Copyright (c) SPSS Inc, 2007 §

Example: Manipulating Output: Merging Tables § tables. py module on Developer Central can merge two tables into one. E. g. , Ctables significance tests into main tables § Merge or replace cells with cells from a different table § Flexibly define the join § tables. py can also censor cells, e. g. , blank statistics based on small counts. § Merge example: data on importance of education qualifications for immigration by region of Europe § CTABLES /TABLE qfimedu. Bin BY Region /TITLES TITLE='Qualifications for Immigration' /COMPARETEST TYPE=PROP Copyright (c) SPSS Inc, 2007 §

Ctables Output Copyright (c) SPSS Inc, 2007

Program to Merge BEGIN PROGRAM. import spss, tables cmd=r"""CTABLES /TABLE qfimedu. Bin BY Region /TITLES TITLE='Qualifications for Immigration' /COMPARETEST TYPE=PROP""" tables. merge. Latest(cmd, autofit=False) END PROGRAM. Runs Ctables and merges test table into main table § Using default merge behavior § "If it really is this simple this will generate a lot of excitement for us. " § "This is really fantastic. " Copyright (c) SPSS Inc, 2007 §

Merged Output Copyright (c) SPSS Inc, 2007

Approaches to Creating New Procedures § You can extend SPSS capabilities by building new procedures § § Or use ones that others have built Combine SPSS procedures and transformations with Python logic Poisson regression (SPSS 14) example using iterated CNLR § New raking procedure built over GENLOG § Calculate data aggregates in SPSS and pass to algorithm coded in Python § § Raking procedure starts with AGGREGATE; uses GENLOG Acquire case data and compute in Python Use Python standard modules and third-party additions § Partial Least Squares Regression (pls module) § Copyright (c) SPSS Inc, 2007 § GENLIN in SPSS 15

Adapt Existing Code Libraries § Common to adapt existing libraries or code for use as Python extension modules § § C, C++, VB, Fortran, . . . Python tools and API's to assist Chap 25 in Python in a Nutshell § Tutorial on extending and embedding the Python interpreter § Call R programs with SPSS 16 Copyright (c) SPSS Inc, 2007 §

Partial Least Squares Regression with large number of predictors (even k > N) § Similar to Principal Components but considers dependent variable simultaneously § Calculates principal components of (y, X) then use regression on the scores instead of original data § Equivalent to ordinary regression when number of factors equals number of predictors and one y variable § For more information see An Optimization Perspective on Kernel Partial Least Squares Regression. pdf. Copyright (c) SPSS Inc, 2007 §

The pls Module for SPSS 15 § Strategy Fetches data from SPSS § Uses scipy matrix operations to compute results § § § Third-party module from Cheeseshop Writes pivot tables to SPSS Viewer § Saves predicted values to active dataset Copyright (c) SPSS Inc, 2007 Subject to OMS § SPSS 14 viewer module created pivot table using OLE automation § SPSS 15 has direct pivot table API's §

pls Example: REGRESSION vs PLS GET FILE="c: /spss 15/tutorial/sample_files/car_sales. sav". REGRESSION /STATISTICS COEFF R /DEPENDENT sales /METHOD=ENTER curb_wgt engine_s fuel_cap horsepow length mpg price resale type wheelbas width. begin program. import spss, pls § plsproc defaults to five factors Copyright (c) SPSS Inc, 2007 plsproc("sales", """curb_wgt engine_s fuel_cap horsepow length mpg price resale type wheelbas width""", yhat="predsales") end program.

Results PLS with 5 factors almost equals regression with 11 variables Copyright (c) SPSS Inc, 2007 §

SPSS 16 User Procedures User procedures can be written in Python but specified using SPSS traditional syntax § User never writes or sees Python code § Used as if a built-in SPSS command § EXTENSION command defines command to SPSS via simple XML file § Python module called with syntax already checked and processed by SPSS § More general PLS module § § PLS y 1 y 2 y 3 BY fac 1 fac 2 WITH z 1 z 2 z 3 /CRITERIA LATENTFACTORS=2. Dialog box interface tools in SPSS 17 § In the meantime, use wx. Python or something similar Copyright (c) SPSS Inc, 2007 §

Raking Sample Weights § "Raking" adjusts sample weights to control totals in n dimensions § Example: data classified by age and sex with known population totals or proportions § Calculated by fitting a main effects loglinear model § Not directly available in SPSS Copyright (c) SPSS Inc, 2007 Various adjustments required § Not a complete solution to reweighting §

Raking Module § Strategy: combine SPSS procedures with Python logic § rake. py (from SPSS Developer Central) § § § rake("age sex", [{0: 1140, 1: 1140}, {0: 104. 6, 1: 2175. 4}], finalweight="finalwt") Copyright (c) SPSS Inc, 2007 § Aggregates data via AGGREGATE to new dataset Creates new variable with control totals Applies GENLOG, saving predicted counts Adjusts predicted counts Matches back into original dataset § Does not use MATCH FILES or require a SORT command Written in one (long) day

Extending SPSS Transformations § SPSS 14 programmability can wrap SPSS syntax in Python logic, e. g. , generate COMPUTE commands on the fly § SPSS 15 programmability can Generate new variables directly § Add new cases directly § Create new datasets from scratch § § SPSS 16 has additional dataset capabilities Copyright (c) SPSS Inc, 2007 § Useful when definitions can be expressed in SPSS syntax

trans and extended. Transforms Modules § trans module facilitates plugging in Python code to iterate over cases § Runs as an SPSS procedure Passes the data § Adds variables to the SPSS variable dictionary § Can apply any calculation casewise § Use with Standard Python functions (e. g. , math module) § Any user-written functions or appropriate classes § Functions in extended. Transforms module § Copyright (c) SPSS Inc, 2007 §

trans and extended. Transforms Modules trans strategy § Pass case data through Python code writing result back to SPSS in new variables § extended. Transforms collection of 12 functions to apply to SPSS variables, including Regular expression search/replace § soundex and nysiis functions for phonetic equivalence § Date/time conversions based on patterns § Copyright (c) SPSS Inc, 2007 §

Python Regular Expressions § Pattern matching in text strings § If you use SPSS index or replace, you need these § Standardize string data (Mr, Mr. , Herr, Senor, . . . ) § Extract data from loosely structured text "simvastatin-- PO 80 mg TAB" -> "simvastatin", "80" § Patterns can be simple strings (as with SPSS index) or complex patterns § Pick out variable names with common parts § Can greatly simplify code Copyright (c) SPSS Inc, 2007 §

Write to Me! Copyright (c) SPSS Inc, 2007