smovie: Some movies to teach statistical concepts 1.1.6

smovie 1.1.6

Bug fixes

• Fixed issues with the incorrect use of \itemize in some Rd files.

smovie: Some movies to teach statistical concepts

smovie 1.1.5

Bug fixes and minor improvements

• Corrected the link to cltq() in the main smovie help page.

• Create the help file for the package correctly, with alias smovie-package.

• Activated 3rd edition of the testthat package

smovie: Some movies to teach statistical concepts

smovie 1.1.4

Bug fixes and minor improvements

• The BWidget Tcl/Tk module is no longer a system requirement for the rpanel package on which smovie depends. rpanel checks whether BWidget is available wen it is required and provides instruction if it is not. This task has been removed from smovie.

• In help files and vignettes DOIs are used rather than URLs.

• A typo in a section heading in the vignette has been corrected: "simply" becomes "a simple".

• LazyData: TRUE has been removed from DESCRIPTION, because there is no data/ directory in smovie.

smovie: Some movies to teach statistical concepts

smovie 1.1.3

UCL Eugenics Inquiry

Explicit references to (Karl) Pearson have removed from the package following the Inquiry into the History of Eugenics at UCL.

Bug fixes and minor improvements

• Errors in test name strings supplied to testthat::test_that() have been corrected, to avoid CRAN package check ERRORs.

• In wws() the gradient of the log-likelihood was unnecessarily labelled with a red "score" when the null value of theta0 is less than the MLE of theta. This has been removed.

• In wws(), when model = "norm" the value of mu could not be passed as described. This has been corrected.

• In shypo() there was a bug that meant that the plots did not behave as intended when mu0 is not equal to (the default value of) 0. This has been corrected.

• In the documentation for shypo() a minor typo at the end of the Details section has been corrected: "based on the current value of n" has been deleted.

• In discrete() radio buttons have been added to switch between the version of the the geometric distribution based on the number of trials up to including the first success and the number of failures until the first success.

• In discrete() a typo has been corrected in the 4th line of the documentation of the argument distn: "hypergeometric" should have read "negative binomial".

smovie: Some movies to teach statistical concepts

smovie 1.1.2

Dependencies

• The packages revdbayes and SuppDists have been demoted from Imports to Suggests. An error is thrown if actions that require either of these packages occurs and that packae is not available.

smovie: Some Movies to Illustrate Concepts in Statistics

smovie 1.1.1

New features

• New movie: cltq(). Central Limit Theorem for sample quantiles. Illustrates the sampling distribution of the sample quantiles and the central limit theorem for sample quantiles.

• New movie: mean_vs_median(). Compares the sampling distributions of the mean and median for random samples from either a standard normal distribution or a standard Student t distribution.

• pkgdown documentation at https://paulnorthrop.github.io/smovie/

Bug fixes and minor improvements

• In ett() the sample size n is now allowed to drop to 1 for those cases in which the normalising constants underlying the calculations are finite, such as the exponential, uniform and generalised Pareto distributions. Otherwise, 2 is the smallest allowable value for n.

• When the Wald, Wilks and Score tests movie wws() is called from the main menu, via movies() a binomial example is used, so that the differences between the test statistics can be seen.

• Adjustments have been made to the wws() movie to ensure that the plot stays completely still during the animations.

smovie: Some Movies to Illustrate Concepts in Statistics

smovie 1.0.1

Bug fixes and minor improvements

• The way that objects modified inside the plotting function are saved has been changed to enable two or more panels produced by the same smovie function to operate simultaneously and independently, even if the call is exactly the same.

• The argument plot_par in the functions discrete and continuous didn't work properly. It does now.

• Both hscale and vscale may be passed to the main menu function movies to scale the size of the graphics panel horizontal and vertically.

• In ett() and clt() the legend position in plots of cdfs has been moved to "topleft" better to avoid overlapping the contents of the plot. The argument leg_cex has been added to enable manual control of the size of the legends.

• That smovie requires the R extension BWidget (via its dependence on the package rpanel) is noted in the README file. If BWidget is not installed then a message is given to explain why the smovie functions don't work.

smovie: Some Movies to Illustrate Concepts in Statistics

Some movies to teach statistical concepts

What does smovie do?

The smovie package provides movies to help students to understand statistical concepts. The rpanel package is used to create interactive plots that move to illustrate key statistical ideas and methods. The movies cover the topics of probability distributions; sampling distributions of the mean (central limit theorem), the maximum (extremal types theorem) and the (Fisher transformation of the) correlation coefficient; simple linear regression; hypothesis testing.

An example

The function wws produces a movie to visualise the Wald, Wilks and score likelihood-based test statistics, for a model with a scalar unknown parameter. The user can change the value of the parameter under a simple null hypothesis and observe the effect on the test statistics. The following code uses the log-likelihood from a binomial experiment and considers the null hypothesis that the success probability θ is equal to θ₀. The user may specify their own log-likelihood.

wws(theta0 = 0.5)

Installation

To get the current released version from CRAN:

install.packages("smovie")

Vignettes

See vignette("smovie-vignette", package = "smovie") for an overview of the package.