Log message:
(finance/R-fBasics) Updated from 3042.89 to 4041.97, make test passed
# fBasics 4041.97
- fixed the missing package qualifier in a link to another ackage.
# fBasics 4032.96
- modified `interactivePlot` to allow also functions in argument
`plotFun`. Previously `plotFun` had to be a character vector containing the
names of the functions. Adjusted the documentation and added an example to
this effect.
- function `.distCheck` is defunct, use `distCheck` instead. To allow
compatibility with older versions of packages that import fBasics (fGarch,
fExtremes), `.distCheck` is exported but new code should not use it.
- removed generic `.print`, its 'ssd' method, and the S3 registration for the
latter, all unused.
- clarified in the documentation for `histPlot`, `densityPlot` and
`logDensityPlot` that argument `x` must be from class "timeSeries" \
(each of
them starts with `stopifnot(is.timeSeries(x))`). Previously it was stated that
argument `x` can be of any class that can be converted to \
"timeSeries" with
`as.timeSeries`.
- contrary to the documentation, the return values from the \code{p/d/q/r*}
functions (such as `dhyp`, `phyp`, `qhyp`) for the generalised distributions
do not contain attribute \code{"param"}. This is now documented \
correctly.
- removed attribute 'control' from the return value of `rhyp` and similar
functions. It was undocumented (or rather, the documentation stated that it is
called "param"; in any case, wouldn't be usable in code).
- argument 'alpha' of `dhyp`, `phyp`, `qhyp` and `rhyp` can now be also a
vector of length 4 containing the four parameters.
- `hypMode` now throws error if argument 'pm' doesn't have one of the allowed
values.
- the return values of `*Mean()`, `*Var()`, `*Skew()`, `*Kurt()`, `*Moments()`,
where '*' is a distribution prefix (such as 'gh' or 'ght'), are now all named.
Previously, only some of them were.
- removed a number of inconsistencies in the documentation.
# fBasics 4031.95
- now `characterTable` doesn't try to print/plot invalid non-ASCII UTF8
characters (fixes CRAN issue from around 2023-10-06).
- `symbolTable` now plots more characters and issues warnings only for
characters in `26:31` which are ignored according to `?points` but are
actually plotted.
- now using default R Random generators in tests, since the old ones were
causing numerous warnings from tests.
- removed deprecated function `listDescription`, use `utils::packageDescription`
instead.
- fixed a bug in `histPlot`, causing a puzzling error when argument
`fit = FALSE`.
# fBasics 4022.94
- `dagotest()` was returning NaN for the value of the test statistic based on
kurtosis when the argument of a cube root was negative. Reported by Cameron
Willden who suggested the fix.
- the name of the first column of dataset `msft.dat` is not mangled any more (it
was "X.Y..m..d" and now is "%Y-%m-%d", as in the input \
file), the rest of the
dataset is as before. The file from which the dataset is created is now taken
from an identical file in package `timeSeries`.
- the `show` methods for classes `fHTEST` and `fDISTFIT` no longer print
'Description:' when slot `description = ""`.
- many fitting functions, e.g., `nFit`, were inadvertently ignoring argument
`description` by using `description = description()` instead of
`description = description` in the call to `new` when creating the return
value.
- `correlationTest`, `pearsonTest`, `spearmanTest`, `kendallTest`, `ks2Test`,
`scaleTest`, `varianceTest`, and `locationTest` now set slot `description` of
the result to `""` by default. Previously they were setting it to \
the current
date/time, with the consequence that the functions gave different objects at
each run. Use `description = date()` in the call if you want the
date. Similarly for `ksnormTest`, `shapiroTest`, `jarqueberaTest`, `dagoTest`,
`normalTest`, `adTest`, `cvmTest`, `lillieTest`, `pchiTest`, `sfTest`,
`jbTest`, `nFit`, `tFit`, `stableFit`, and many others, which had default
setting `description = description()`.
# fBasics 4021.93
- fixed CRAN warnings about a function declaration without a prototype, which is
deprecated by modern C compilers.
- further updates and improvements to the documentation.
- website built with `pkgdown` is now linked to in file `DESCRIPTION`.
# fBasics 4021.92
- new maintainer: Georgi N. Boshnakov.
- in DESCRIPTION, moved `timeDate` and `timeSeries` from `Depends:` to
`Imports:`. This necessitated changes in the examples and unit tests to get
them run smoothly. The implications for end users and developers are discussed
in separate sections below.
- in NAMESPACE, now export selectively rather than with a generic pattern. No
longer export symbols starting with a dot. Reexported some functions that can
reasonably be expected when `fBasics` is attached (the list of such functions
can be adjusted).
- tidied up the documentation somewhat. There is now a `_pkgdown` file with the
functions in the package organised by topic. Run `pkgdown::build_site()` on
the source directory (or unpacked tarball) to build the site locally.
- removed `.HedgeFund1` and `.HedgeFund2` - they were just used to build the
dataset `HedgeFund`.
## Notes for users
The changes are aimed at making the package easier to manage and reduce the side
effects for users from attaching it.
- Since packages `timeDate` and `timeSeries` are no longer attached by
`library(fBasics)`, users may need to attach them explicitly, if they use
functions from them directly. That said, some functions from `timeDate` and
`timeSeries` are reexported by `fBasics`, so most users may not even notice
any change.
- Undocumented functions (mostly starting with a dot, '.') are no longer
exported. If you believe that a non-exported and undocumented function should
be exported, please open a bug report, giving your reasons.
## Notes for developers
The notes for users apply to developers, as well. In addition:
- We continue to export some (undocumented) functions starting with a '.' to
avoid breaking packages that import `fBasics`. Please consider using
documented equivalents, if possible. Alternatively, let us know that they are
useful, so that we can document them.
|