-
Notifications
You must be signed in to change notification settings - Fork 0
/
ChangeLog
131 lines (86 loc) · 4.64 KB
/
ChangeLog
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
2016-12-11 yves <deville.yves@alpestat.com>
* Added `varnames` and `varnames<-` S3 methods for "GPtail".
2016-03-09 yves <deville.yves@alpestat.com>
* Completed many changes in the SeaLevGuide. The .Rnw file must
now be compiled with knitr and no longer with Sweave.
2016-03-08 yves <deville.yves@alpestat.com>
* Fixed bug in 'convSL' (new version). When computing the finite
differences on 'par.y', elements names are lost in 'par.plus[ ,
ip]' when 'p.y' is 1. The function 'myFuns$F.y' could then no
longer work as it should. Solved with an uggly 'if (p.y > 1L)'.
* Fixed bug in 'convSL2': 'threshold.y' was not used for
quadrature points to shift the result.
* Fixed bug in 'GPtail': 'threshold.y' was not used as it should.
* Renamed convSL to 'convSLOld' then 'convSL2' to 'convSL'
2016-03-02 yves <deville.yves@alpestat.com>
* Fixed a bug in 'convSL2'. When the distribution had only one
parameter, the matrix of gradients of S was turned into a vector
due to a missing 'drop = FALSE'.
2016-01-11 yves <deville.yves@alpestat.com>
* Changed the default vale of 'N.quad' in 'convSL2', because the
default was too small which caused bumpy curves on Return level
plots.
2015-11-30 yves <deville.yves@alpestat.com>
* Added the GPtail (S3) class and some method. The GPtail creator
function computes the convolution of a spline density and a GPD
with provided parameters. The computation uses an enterly new
formula.
* Added the SplineDensity (S3) class and some methods.
* Added the new function convSL2, a new implementation of convSL
which should replace it.
2013-07-08 yves <deville.yves@alpestat.com>
* Added a NAMESPACE.
* Vignette 'SeaLevGuide' was slightly modified. The 'fRenouv'
function was replaced in code chunks by 'Renouv' and
the covariance of a "Renouv" object is now found using the 'vcov'
method. The test function \phi(x, y) received its missing 'y'
argument. Some errors concerning english were corrected, e.g.
'conditional to' -> 'conditional on'.
* The .Rbuildignore file was edited in order to ship with the
package only the pdf file, the .Rnw code and R code.
2010-11-16 yves <deville.yves@alpestat.com>
* RSLplot now has args 'below' and 'alpha.below'. The
semi-transparent rectangle is no longer drawn in 'convSL' but in
'RSLplot'. This rectangle can partially occult return
level/confidence lines yet letting empirical points unaffected.
* Typos found by Lise Bardet in the Rd files / manual were
corrected. Howevet the formal 'pct.conf' of convSL was not renamed
to math the names of Renext, since Renext formals names might be
changed in the future.
* Vignette was slightly modified. The names of Pugh and Vassie is
cited at the begining. The example with empirical points was
slightly modified.
2010-11-04 yves <deville.yves@alpestat.com>
* RSLplot now supports the plotting of experimental points
given through 'z' and 'duration' formals.
* convSL has an 'alpha.below' formal allowing the user to "gray"
or partially occult the region below HAT+threshold on the return
level plot.
* Rd files were corrected. Note that the use of references in
french lead to using the \enc{}{} macro and the \encoding{UTF-8}
section. The manual is generated using Rd2dvi with with an
'--encoding=utf8' option, but it seems to works fine without it.
* Vignette was corrected. The plotting of experimental points
is described (how return periods are computed) and illustrated
with examples.
2010-09-17 yves <deville.yves@alpestat.com>
* convSL allows the superposition of an asymptotic return
level curve for exponential and GPD POT distributions,
only for postitive shape in the GPD case.
The exponential case provides a validation for the convolution
comptation, since the asymptotic curve is exact for return levels
> threshold.y + xmax (see Vignette) and is computed independently
from the convolution.
* convSL arguments names percents for confidence bands are now
given in 'pct.conf'. The threshold. in 'threshold.y' an now be set
to 'NA' in which case the distribution of Y is for the surge
itself rather than the surge exceedance Y-u. This allows the use
of GEV distribution as in B. Simon's Book. Graphical parameters
are passed to 'RSLplot' (e.g. 'Tlim' and 'problim') or to plot
('main', 'ylim').
* RSLplot Arguments of the 'RSLplot' function are changed. The
args concerning confidence bands are named ***.conf. Two new ars
'Tlim' and 'problim' allow the user to set the x axis limits in
periods (years) or probability. The default values have been
changed to cover larger ranges. It is possible to use filled
confidence regions, to set the color and line type of these.