On August 4 we asked you â€“ our customers â€“ to tell us what Matrix features you would like to see added to SPSS Statistics. Just 3.5 months later we are excited to let you know that we took your suggestions to heart and have delivered an impressive set of enhancements.
These enhancements are now available to SPSS Statistics Subscribers with an in-software update.
We are sure that you will want to see what these enhancements will look like. Of course, screenshots and text is nice, but nothing says â€˜we listened to youâ€™ as hands-on examples â€“ be sure to check out the short videos under each of the enhancements for a quick taste.
Ability to view matrix output as a pivot table
We have added a MDISPLAY command to global set and show commands â€“ it allows to toggle between TEXT (original way of displaying matrices) and TABLES (a more attractive pivot table that is consistent with outputs of other procedures).
Support for 64-byte names to align with rest of the Statistics platform
Long variable names (up to 64 bytes) are now fully supported for names of matrices and/ vectors when used with any statement.
Improved dataset referencing
To make matters easier and match Statistics procedures, we have implemented support for Dataset references with GET and MGET commands.
44 additional matrix functions â€“ so many and so good, all theyâ€™re missing is sprinkles and a waffle cone!
Because we want to maximize usability and time-to-value, we want to make sure you can get up and running as fast as possible â€“ as such we tried to leverage familiar options as much as possible.
The new functions are added under the same familiar names that are used in the traditional COMPUTE command. With the exception of PROBIT, they follow the same pattern of FUNCTION.DISTRIBUTION (i.e. CDF.BINOM).
The full list of new statistical functions in MATRIX is listed below:
|Matrix Function||Matrix Function||Matrix Function||Matrix Function|
|IDF.CHISQ (prob, df)||IDF.F (prob, df1, df2)||IDF.T (prob, df)||CDF.BINOM (quant, n, prob)|
|PROBIT (prob)||CDF.BETA (quant, shape1, shape2)||IDF.BETA (prob, shape1, shape2)||CDF.BERNOULLI (quant, prob)|
|CDF.GAMMA (quant, shape, rate)||IDF.GAMMA (prob, shape, rate)||CDF.NORMAL (quant, mean, stddev)||IDF.NORM (prob, mean, stddev)|
|CDF.POISSON(quant, mean)||CDF.EXP(quant, scale)||IDF.EXP(prob, scale)||CDF.LOGISTIC(quant, mean, scale)|
|IDF.LOGISTIC(prob, mean, scale)||CDF.IGAUSS(quant, loc, scale)||IDF.IGAUSS(prob, loc, scale)||CDF.NEGBIN(quant, thresh, prob)|
|CDF.UNIFORM(quant, min, max)||IDF.UNIFORM(prob, min, max)||CDF.WEIBULL(quant, a, b)||IDF.WEIBULL(prob, a, b)|
|CDF.BVNOR(quant1, quant2, corr)||CDF.GEOM(quant, prob)||CDF.LNORMAL(quant, a, b)||IDF.LNORMAL(prob, a, b)|
|CDF.LAPLACE(quant, mean, scale)||IDF.LAPLACE(prob, mean, scale)||CDF.PARETO(quant, threshold, shape)||IDF.PARETO(prob, threshold, shape)|
|NCDF.CHISQ(quant, if, nc)||CDF.CAUCHY(quant, loc, scale)||IDF.CAUCHY(prob, loc, scale)||CDF.HYPER(quant, total, sample, hits)|
|CDF.SRANGE(quant, a, b)||IDF.SRANGE(prob, a, b)||CDF.HALFNRM(quant, mean, stddev)||IDF.HALFNRM(prob, mean, stddev)|
|CDF.SMOD(quant, a, b)||IDF.SMOD(prob, a, b)||NCDF.F(quant, df1, df2, nc)||NCDF.T(quant, df, nc)|
Some other exciting enhancements coming with this update are:
– Reduced startup time of Subscription by up to 3X.
– Added a direct link to product support in the Help tab.
– Python extensions no longer installed by default (however, Python Essentials still comes installed by default). You can download extensions from the Extension Hub. Check out an excellent blog post on how to do that with an exciting new STATS Power extension.
The Statistics trial has been updated to include these features and all paid subscribers will receive an update.
For more details please be sure to review the documentation and do not hesitate to reach out!