### 1. Highlights

● produced in 2007

● 11 new directives and 34 new procedures

● fit regressions, curves and nonlinear models when the residuals have an AR1 structure or have correlations defined by a power-distance model (`RAR1`

and `NLAR1`

)

● fit a quadratic surface and estimate its stationary point (`RQUADRATIC`

)

● circular regression (`RCIRCULAR`

)

● meta analysis to form fixed or random combined estimates using results saved from individual trials (`META`

)

● Cochran-Mantel-Haenszel test (`CMHTEST`

)

● no limit on the number of comparison contrasts fitted by `ANOVA`

, regression contrasts limited only by the number of d.f. of the factor whose contrasts are being calculated

● F statistics for fixed terms in `REML`

● produce experimental designs efficient under analysis of covariance (`COVDESIGN`

)

● hierarchical generalized *nonlinear* models (`HGNONLINEAR`

)

● negative-binomial distribution and logratio link in generalized estimating equations (`GEE`

) and generalized linear mixed models (`GLMM`

)

● estimate parameters of beta binomial distribution (`BBINOMIAL`

)

● estimate false discovery rate by Bonferroni-type procedure (`FDRBONFERRONI`

)

● MCMC analyses by running the WinBUGS or OpenBUGS systems from GenStat (`BGXGENSTAT`

)

● redundancy analysis (`RDA`

)

● species accumulation curves (`ECACCUMULATION`

)

● Shapiro-Wilk test for Normality (`WSTATISTIC`

)

● least significant intervals (`SEDLSI`

)

● enhanced text manipulation: break a text up into individual words (`TXBREAK`

), construct a text by apppending or concatenating values from other variates, factors, texts or pointers (`TXCONSTRUCT`

), search for a sub-text within a text (`TXFIND`

), or search for strings in the lines of a text (`TXPOSITION`

)

● form multiple-response factors from free-response data (`FFREERESPONSEFACTOR`

)

● tabulation of skewness and kurtosis coefficients (`TABULATE`

)

● generalized linear models and bootstrap analysis of survey data (`SVBOOT`

and `SVGLM`

)

● control over which margins to print of tables (`PRINT`

)

● plots of one- or two-way tables of means from `REML`

(`VGRAPH`

)

● enhanced graphics customization: outlines of bar charts (`BARCHART`

), histograms (`DHISTOGRAM`

) and pie charts (`DPIE`

); annotation of contour lines by their heights (`DCONTOUR`

); boxes around keys and surfaces (`FRAME`

); styles of grids in shade plots (`DSHADE`

); shading of the sides of surfaces (`DSURFACE`

); more control over the positions of the x-, y- and z-axes, and of the sizes of the margins between the ends of the axes and the lowest/highest plotted points (`XAXIS`

, `YAXIS`

and `ZAXIS`

)

● add annotation to graphs (`DTEXT`

)

● create, save and reload custom graphics environments (`DLOAD`

and `DSAVE`

)

● define 256 different colours for graphics (previously 64; see `COLOUR`

)

● define 256 different graphics pens (previously 64; see `PEN`

)

### 2. What’s new

**2.1 Directives**

`BARCHART`

plots bar charts in high-resolution graphics.

`CALLS`

lists library procedures called by a procedure.

`DLOAD`

loads the graphics environment settings from an external file.

`DSAVE`

saves the current graphics environment settings to an external file.

`LPCONTOUR`

produces contour maps of two-way arrays of numbers using character (i.e. line-printer) graphics. This is a synonym of the existing `CONTOUR`

directive, which may be modified to use high-resolution graphics in a future release.

`LPGRAPH`

produces point and line plots using character (i.e. line-printer) graphics. This is a synonym of the existing `GRAPH`

directive, which may be modified to use high-resolution graphics in a future release.

`LPHISTOGRAM`

produces histograms using character (i.e. line-printer) graphics. This is a synonym of the existing `HISTOGRAM`

directive, which may be modified to use high-resolution graphics in a future release.

`TXBREAK`

breaks up a text structure into individual words.

`TXCONSTRUCT`

forms a text structure by appending or concatenating values of scalars, variates, texts, factors or pointers; allows the case of letters to be changed or values to truncated and reversed.

`TXFIND`

finds a subtext within a text structure.

`TXPOSITION`

locates strings within the lines of a text structure.

**2.2 Procedures**

`AMCOMPARISON`

performs pairwise multiple comparison tests for `ANOVA`

means.

`BBINOMIAL`

estimates the parameters of the beta binomial distribution.

`BGIMPORT`

imports MCMC output in CODA format produced by WinBUGS or OpenBUGS.

`BGPLOT`

produces plots for output and diagnostics from MCMC simulations.

`BGXGENSTAT`

runs WinBUGS from GenStat in batch mode using scripts.

`CMHTEST`

performs the Cochran-Mantel-Haenszel test.

`COVDESIGN`

produces experimental designs efficient under analysis of covariance.

`DCOVARIOGRAM`

plots 2-dimensional auto- and cross-variograms.

`DOTHISTOGRAM`

plots dot histograms.

`DTEXT`

adds text to a graph.

`DVARIOGRAM`

plots fitted models to an experimental variogram.

`ECACCUMULATION`

plots species accumulation curves for samples or individuals.

`FACLEVSTANDARDIZE`

redefines a list of factors so that they have the same levels or labels.

`FCORRELATION`

forms the correlation matrix for a list of variates.

`FDRBONFERRONI`

estimates false discovery rates by a Bonferroni-type procedure.

`FFREERESPONSEFACTOR`

forms multiple-response factors from free-response data.

`GETNAME`

forms the name of a structure according to its `IPRINT`

attribute.

`HGNONLINEAR`

defines nonlinear parameters for the fixed model of a hierarchical generalized linear model.

`LSIPLOT`

plots least significant intervals, saved from `SEDLSI`

.

`META`

combines estimates from individual trials.

`MIN1DIMENSION`

finds the minimum of a function in one dimension.

`NLAR1`

fits curves with an AR1 or a power-distance correlation model.

`PDUPLICATE`

duplicates a pointer, with all its components.

`PTROTATE`

rotates a point pattern.

`RAR1`

fits regressions with an AR1 or a power-distance correlation model.

`RCIRCULAR`

does circular regression of mean direction for an angular response.

`RDA`

performs redundancy analysis.

`RQUADRATIC`

fits a quadratic surface and estimates its stationary point.

`SAGRAPES`

produces statistics and graphs for checking sensory panel performance.

`SEDLSI`

calculates least significant intervals.

`SVBOOT`

bootstraps data from random surveys.

`SVGLM`

fits generalized linear models to survey data.

`VGRAPH`

plots one- or two-way tables of means from `REML`

.

`WSTATISTIC`

calculates the Shapiro-Wilk test for Normality.

### 3. What’s changed

Most of the changes are completely compatible with Release 9, the previous release. There are a few commands, however, where new options or parameters have been inserted into the existing lists. These may cause problems in statements where option or parameter names have been omitted or abbreviated (see Section 1.7.1 of Part 1 of the *Guide to the GenStat Command Language* for details). To avoid any difficulty, the name of the option/parameter after the new option/parameter should be given explicitly, and not abbreviated to fewer than four characters.

Any command, where changes in Release 10 may cause incompatibilities in existing programs, is marked in Sections 3.1 and 3.2 by the symbol ^{†}. The full details are given in Section 3.3.

**3.1 Directives**

`ANOVA`

now has no limit on the number of comparisons that can be made between levels of a factor using the `COMPARISON`

function.

^{†}`DCONTOUR`

can now annotate the contours by their heights (as an alternative to an integer indexing a list of contour heights in the key).

^{†}`DHISTOGRAM`

allows you to control the use and appearance of the lines drawing the outlines around the bars or the perimeter of the plot.

^{†}`DPIE`

allows you to control the use and appearance of the lines drawing the outlines around the slices or the perimeter of the pie chart. It also allows you to control how the slices are annotated in the description.

^{†}`DSURFACE`

allows you to shade the sides of the surface.

^{†}`DSHADE`

allows to to control the way in which the grid is drawn.

`FCLASSIFICATION`

can now reorder the terms in the model; in particular, this means that a term that is marginal to an earlier term will be moved to come before it, rather than being deleted.

`FITCURVE`

allows for weight matrices, provided these do not define non-zero covariances between groups, if any are defined (by a factor in the model).

^{†}`FRAME`

can control the sizes of boxes to draw around keys, and the position of the graph title.

`PREDICT`

allows classifying factors to be specified to run in parallel with other classifying factors (as in `VPREDICT`

).

`PRINT`

allows you to print only some of the margins of tables, can now omit lines of tables that contain only missing values.

^{†}`REML`

and `VDISPLAY`

have a new option `FMETHOD`

to control the production of F-statistics for fixed terms.

^{†}`RKEEP`

has an option `IGNOREFAILURE`

that allows you to save information from generalized linear models that have not have fitted successfully.

`TABULATE`

can form tables of skewness and kurtosis coefficients, and of standard errors of means.

`TERMS`

can print monitoring information from the fitting of the null model that contains just a constant.

^{†}`XAXIS`

, ^{†}`YAXIS`

and ^{†}`ZAXIS`

provide more control over the positions of the x-, y- and z-axes, and of the sizes of the margins between the ends of the axes and the lowest/highest plotted points.

**3.2 Procedures**

`A2DISPLAY`

, ^{†}`A2WAY`

and ^{†}`A2KEEP`

have a new option `RMETHOD`

to control the type of residuals (simple or standardized) that are formed.

can plot residuals from any stratum (error term), not just from the final residual term, and the user can supply an overall title (to use instead of the identifier of the y-variate).

`AFIELDRESIDUALS`

can now produce shade plots and can now plot standardized residuals from the final stratum (as an alternative to simple residuals).

^{†}`AGRAPH`

can now plot bar charts, and also now allows e.s.e.’s or l.s.d.’s to be used instead of s.e.d.’s.

`AMTDISPLAY`

and ^{†}`AMTIER`

now allow the choice of whether to combine sets of pseudo-terms in the analysis-of-variance table or present them in separate lines.

^{†}`AMTIER`

now allows you to constrain the type of balance within the second and third structure formulae.

^{†}`AREPMEASURES`

allows numbers or labels to be specified to identify the time points in the output.

^{†}`A2KEEP`

has a new option `RMETHOD`

to control the type of residuals (simple or standardized) that are saved.

`CENSOR`

has a new option `MAXCYCLE`

to control the number of iterations.

`CVAPLOT`

has new options `CREGION`

and `CIPROBABILITY`

to control the type and size of confidence region.

^{†}`GEE`

now allows the negative-binomial distribution and logratio link.

^{†}`GLMM`

now allows the negative-binomial distribution and logratio link.

^{†}`HGFIXEDMODEL`

has a new option `DLINK`

to allow a reciprocal link function to be selected instead of the log link for the dispersion model for phi, and a new option `DOFFSET`

to include an offset there.

^{†}`HGPLOT`

has a new option `DHGRANDOMTERM`

to plots to be produced for random model terms in the HGLM modelling the dispersion in a DHGLM.

^{†}`HGRANDOMMODEL`

has a new parameter `DLINK`

to allow a reciprocal link function to be selected instead of the log link for the dispersion models for the random terms, and a new option `DOFFSET`

to include offsets there.

`MAANOVA`

now allows the limits for contrasts and deviations to be changed from the default (`ANOVA`

) values.

`MTABULATE`

can form tables of standard deviations, skewness and kurtosis coefficients, and of standard errors of means.

`PARTIALCORRELATIONS`

can now calculate weighted correlations.

`TTEST`

can now calculate weighted t-tests.

**3.3 Incompatibilities**

`A2WAY` procedure |
option `RMETHOD` inserted before `EXIT` . |
---|---|

`A2KEEP` procedure |
option `RMETHOD` inserted before `EXIT` . |

`AGRAPH` procedure |
options `PSE` and `LSDLEVEL` inserted before `SAVE` . |

`AMTIER` procedure |
options `F2BALANCE` and `F3BALANCE` inserted before `PSEUDOTERMS` . |

`AREPMEASURES` procedure |
`TIMEPOINTS` option inserted before `FPROBABILITY` option. |

`AUKEEP` procedure |
option `RMETHOD` inserted before `SAVE` . |

`CONTOUR` directive |
may be replaced by a high-resolution graphics version in a future release. You may wish to switch now to the `LPCONTOUR` directive (which has the same syntax) to avoid future incompatibilities. |

`DCONTOUR` |
option `ANNOTATION` inserted before `SCREEN` . |

`DHISTOGRAM` |
options `OUTLINE` and `PENOUTLINE` inserted before `SCREEN` . |

`DPIE` |
options `ANNOTATION` , `OUTLINE` and `PENOUTLINE` inserted before `SCREEN` . |

`DSURFACE` |
option `PENSIDE` inserted before `NCONTOURS` . |

`DSHADE` |
option `PENGRID` inserted before `SCREEN` . |

`FRAME` directive |
options `BOXSURFACE` and `BOXKEY` inserted before `PENTITLE` , and option `TPOSITION` inserted before `SAVE` . |

`GEE` procedure |
options `AGGREGATION` and `KLOGRATIO` inserted before `QUADESTIMATION` . |

`GLMM` procedure |
options `AGGREGATION` and `KLOGRATIO` inserted before `OWNDIST` . |

`GRAPH` directive |
may be replaced by a high-resolution graphics version in a future release. You may wish to switch now to the `LPGRAPH` directive (which has the same syntax) to avoid future incompatibilities. |

`HGFIXEDMODEL` procedure |
option `DLINK` inserted before `DTERMS` . |

`HGPLOT` procedure |
parameter `DHGRANDOMTERM` inserted before `RMETHOD` . |

`HGRANDOMMODEL` procedure |
parameter `DLINK` inserted before `DFORMULA` , and parameter `DOFFSET` inserted before `LMATRIX` . |

`HISTOGRAM` directive |
may be replaced by a high-resolution graphics version in a future release. You may wish to switch now to the `LPHISTOGRAM` directive (which has the same syntax) to avoid future incompatibilities. |

`REML` directive |
option `FMETHOD` inserted before `WORKSPACE` . |

`RKEEP` directive |
option `IGNOREFAILURE` inserted before `SAVE` . |

`XAXIS` directive |
options `MLOWER%` , `MUPPER%` , `DECIMALS` , `YOMETHOD` and `ZOMETHOD` inserted before `SAVE` . |

`YAXIS` directive |
options `MLOWER%` , `MUPPER%` , `DECIMALS` , `XOMETHOD` and `ZOMETHOD` inserted before `SAVE` . |

`ZAXIS` directive |
options `MLOWER%` , `MUPPER%` , `DECIMALS` , `XOMETHOD` and `YOMETHOD` inserted before `SAVE` . |