P-STAT Advanced Statistics [P-Stat Inc.]

P-STAT's Advanced Statistics module features canonical correlation,nonlinear regression, survival analysis, and time series.Canonical Correlation: The Canonical.Cor command analyzes therelationships between two sets of variables by finding a linearcombination of variables in one set that has maximum correlation with alinear combination in the other set. Multiple pairs of linearcombinations are sought. The canonical correlations and theirsignificance are reported. Output files of canonical coefficients(weights) and correlations (structure coefficients), correlations of theinput variables, and canonical variables (variates or scores) may berequested.NonLinear Regression: The NL.Regression command estimates the parametersfor nonlinear regression models using the Rosenbrock optimizationprocedure. Other nonlinear functions may be optimized as well. The sumof the squared residuals is minimized, producing least-squares estimates,but any loss function (objective function) may be minimized.Survival Analysis: P-STAT has two commands for survival analysis: TheLIFE.TABLE command analyzes lifetable (survival time) data using thelifetable (i.e. actuarial) method. The PRODUCT.LIMIT command calculatesthe survival distribution function using the product-limit estimator.Time Series: The ARIMA command fits and forecasts univariate time seriesdata using the autoregressive integrated moving average model of Box andJenkins. The model is a linear combination of past data values and pasterrors (random chock). The time series data are ordered measurements,equally spaced in time, without any missing values. Conditional leastsquares or maximum likelihood estimation are used to calculate theparameters in the model.The Advanced Statistics module requires the Base P-STAT Program.

Language: Fortran and C
Source Avail: No
Product Special Handling: None.
Operating Systems: Solaris Sparc 1.0,2.0,2.1 Solaris x86 2.0

P-Stat Inc.
230 Lambertville Hopewell Rd Road
Hopewell, NJ 08525
USA
Phone: (609) 466-9200
Fax: (609) 466-1688

Data Analysis

P-STAT Advanced Statistics

P-Stat Inc.