root/doc/latex/classbdm_1_1migamma.tex @ 261

Revision 261, 9.0 kB (checked in by smidl, 16 years ago)

doc

Line 
1\hypertarget{classbdm_1_1migamma}{
2\section{bdm::migamma Class Reference}
3\label{classbdm_1_1migamma}\index{bdm::migamma@{bdm::migamma}}
4}
5Inverse-Gamma random walk. 
6
7
8{\tt \#include $<$libEF.h$>$}
9
10Inheritance diagram for bdm::migamma:\nopagebreak
11\begin{figure}[H]
12\begin{center}
13\leavevmode
14\includegraphics[width=76pt]{classbdm_1_1migamma__inherit__graph}
15\end{center}
16\end{figure}
17Collaboration diagram for bdm::migamma:\nopagebreak
18\begin{figure}[H]
19\begin{center}
20\leavevmode
21\includegraphics[height=400pt]{classbdm_1_1migamma__coll__graph}
22\end{center}
23\end{figure}
24\subsection*{Public Member Functions}
25\begin{CompactItemize}
26\item 
27\hypertarget{classbdm_1_1migamma_07c5970da0e578ce8a428f1ebf46a459}{
28\hyperlink{classbdm_1_1migamma_07c5970da0e578ce8a428f1ebf46a459}{migamma} (const \hyperlink{classbdm_1_1RV}{RV} \&\hyperlink{classbdm_1_1mpdf_9bcfb45435d30983f436d41c298cbb51}{rv}, const \hyperlink{classbdm_1_1RV}{RV} \&\hyperlink{classbdm_1_1mpdf_5a5f08950daa08b85b01ddf4e1c36288}{rvc})}
29\label{classbdm_1_1migamma_07c5970da0e578ce8a428f1ebf46a459}
30
31\begin{CompactList}\small\item\em Constructor. \item\end{CompactList}\item 
32\hypertarget{classbdm_1_1migamma_1d7023b1565551d0260eb1ba832bebaf}{
33void \hyperlink{classbdm_1_1migamma_1d7023b1565551d0260eb1ba832bebaf}{set\_\-parameters} (double k0)}
34\label{classbdm_1_1migamma_1d7023b1565551d0260eb1ba832bebaf}
35
36\begin{CompactList}\small\item\em Set value of {\tt k}. \item\end{CompactList}\item 
37\hypertarget{classbdm_1_1migamma_7a34b1e2e3aa2250d7c0ed7df1665b8c}{
38void \hyperlink{classbdm_1_1migamma_7a34b1e2e3aa2250d7c0ed7df1665b8c}{condition} (const vec \&val)}
39\label{classbdm_1_1migamma_7a34b1e2e3aa2250d7c0ed7df1665b8c}
40
41\begin{CompactList}\small\item\em Update {\tt ep} so that it represents this \hyperlink{classbdm_1_1mpdf}{mpdf} conditioned on {\tt rvc} = cond. \item\end{CompactList}\item 
42virtual vec \hyperlink{classbdm_1_1mpdf_e4848a428d8ef0549c6e4a9ed386d9f2}{samplecond} (const vec \&cond, double \&ll)
43\begin{CompactList}\small\item\em Returns a sample from the density conditioned on {\tt cond}, $x \sim epdf(rv|cond)$. \item\end{CompactList}\item 
44virtual mat \hyperlink{classbdm_1_1mpdf_ee26963a637b2ea1fb1933652981e652}{samplecond\_\-m} (const vec \&cond, vec \&ll, int N)
45\begin{CompactList}\small\item\em Returns. \item\end{CompactList}\item 
46\hypertarget{classbdm_1_1mpdf_6336a8a72462e2a56a3989a220f18b1b}{
47virtual double \hyperlink{classbdm_1_1mpdf_6336a8a72462e2a56a3989a220f18b1b}{evallogcond} (const vec \&dt, const vec \&cond)}
48\label{classbdm_1_1mpdf_6336a8a72462e2a56a3989a220f18b1b}
49
50\begin{CompactList}\small\item\em Shortcut for conditioning and evaluation of the internal \hyperlink{classbdm_1_1epdf}{epdf}. In some cases, this operation can be implemented efficiently. \item\end{CompactList}\item 
51\hypertarget{classbdm_1_1mpdf_0b0ed1ed663071bb7cf4a1349eb94fcb}{
52virtual vec \hyperlink{classbdm_1_1mpdf_0b0ed1ed663071bb7cf4a1349eb94fcb}{evallogcond\_\-m} (const mat \&Dt, const vec \&cond)}
53\label{classbdm_1_1mpdf_0b0ed1ed663071bb7cf4a1349eb94fcb}
54
55\begin{CompactList}\small\item\em Matrix version of evallogcond. \item\end{CompactList}\item 
56\hypertarget{classbdm_1_1mpdf_b3aba7311038bf990d706a64cab60cf8}{
57\hyperlink{classbdm_1_1RV}{RV} \hyperlink{classbdm_1_1mpdf_b3aba7311038bf990d706a64cab60cf8}{\_\-rvc} () const }
58\label{classbdm_1_1mpdf_b3aba7311038bf990d706a64cab60cf8}
59
60\begin{CompactList}\small\item\em access function \item\end{CompactList}\item 
61\hypertarget{classbdm_1_1mpdf_222d5280e309c5a053ba73841e98c151}{
62\hyperlink{classbdm_1_1RV}{RV} \hyperlink{classbdm_1_1mpdf_222d5280e309c5a053ba73841e98c151}{\_\-rv} () const }
63\label{classbdm_1_1mpdf_222d5280e309c5a053ba73841e98c151}
64
65\begin{CompactList}\small\item\em access function \item\end{CompactList}\item 
66\hypertarget{classbdm_1_1mpdf_1892fe3933488942253679f068e9e7f6}{
67\hyperlink{classbdm_1_1epdf}{epdf} \& \hyperlink{classbdm_1_1mpdf_1892fe3933488942253679f068e9e7f6}{\_\-epdf} ()}
68\label{classbdm_1_1mpdf_1892fe3933488942253679f068e9e7f6}
69
70\begin{CompactList}\small\item\em access function \item\end{CompactList}\item 
71\hypertarget{classbdm_1_1mpdf_05e843fd11c410a99dad2b88c55aca80}{
72\hyperlink{classbdm_1_1epdf}{epdf} $\ast$ \hyperlink{classbdm_1_1mpdf_05e843fd11c410a99dad2b88c55aca80}{\_\-e} ()}
73\label{classbdm_1_1mpdf_05e843fd11c410a99dad2b88c55aca80}
74
75\begin{CompactList}\small\item\em access function \item\end{CompactList}\end{CompactItemize}
76\subsection*{Protected Attributes}
77\begin{CompactItemize}
78\item 
79\hypertarget{classbdm_1_1migamma_a31b39d4179551b593c9e0d7d756783a}{
80\hyperlink{classbdm_1_1eigamma}{eigamma} \hyperlink{classbdm_1_1migamma_a31b39d4179551b593c9e0d7d756783a}{epdf}}
81\label{classbdm_1_1migamma_a31b39d4179551b593c9e0d7d756783a}
82
83\begin{CompactList}\small\item\em Internal \hyperlink{classbdm_1_1epdf}{epdf} that arise by conditioning on {\tt rvc}. \item\end{CompactList}\item 
84\hypertarget{classbdm_1_1migamma_dc56bc9da542e0103ec16b9be8e5e38c}{
85double \hyperlink{classbdm_1_1migamma_dc56bc9da542e0103ec16b9be8e5e38c}{k}}
86\label{classbdm_1_1migamma_dc56bc9da542e0103ec16b9be8e5e38c}
87
88\begin{CompactList}\small\item\em Constant $k$. \item\end{CompactList}\item 
89\hypertarget{classbdm_1_1migamma_4825c0ef11a148bad9b802a496f56f96}{
90vec $\ast$ \hyperlink{classbdm_1_1migamma_4825c0ef11a148bad9b802a496f56f96}{\_\-beta}}
91\label{classbdm_1_1migamma_4825c0ef11a148bad9b802a496f56f96}
92
93\begin{CompactList}\small\item\em cache of epdf.beta \item\end{CompactList}\item 
94\hypertarget{classbdm_1_1migamma_b6c265b132ff79963bf51dff4c3ef252}{
95vec $\ast$ \hyperlink{classbdm_1_1migamma_b6c265b132ff79963bf51dff4c3ef252}{\_\-alpha}}
96\label{classbdm_1_1migamma_b6c265b132ff79963bf51dff4c3ef252}
97
98\begin{CompactList}\small\item\em chaceh of epdf.alpha \item\end{CompactList}\item 
99\hypertarget{classbdm_1_1mpdf_9bcfb45435d30983f436d41c298cbb51}{
100\hyperlink{classbdm_1_1RV}{RV} \hyperlink{classbdm_1_1mpdf_9bcfb45435d30983f436d41c298cbb51}{rv}}
101\label{classbdm_1_1mpdf_9bcfb45435d30983f436d41c298cbb51}
102
103\begin{CompactList}\small\item\em modeled random variable \item\end{CompactList}\item 
104\hypertarget{classbdm_1_1mpdf_5a5f08950daa08b85b01ddf4e1c36288}{
105\hyperlink{classbdm_1_1RV}{RV} \hyperlink{classbdm_1_1mpdf_5a5f08950daa08b85b01ddf4e1c36288}{rvc}}
106\label{classbdm_1_1mpdf_5a5f08950daa08b85b01ddf4e1c36288}
107
108\begin{CompactList}\small\item\em random variable in condition \item\end{CompactList}\item 
109\hypertarget{classbdm_1_1mpdf_5eea43c56d38e4441bfb30270db949c0}{
110\hyperlink{classbdm_1_1epdf}{epdf} $\ast$ \hyperlink{classbdm_1_1mpdf_5eea43c56d38e4441bfb30270db949c0}{ep}}
111\label{classbdm_1_1mpdf_5eea43c56d38e4441bfb30270db949c0}
112
113\begin{CompactList}\small\item\em pointer to internal \hyperlink{classbdm_1_1epdf}{epdf} \item\end{CompactList}\end{CompactItemize}
114
115
116\subsection{Detailed Description}
117Inverse-Gamma random walk.
118
119Mean value, $\mu$, of this density is given by {\tt rvc} . Standard deviation of the random walk is proportional to one $k$-th the mean. This is achieved by setting $\alpha=\mu/k+2$ and $\beta=\mu(\alpha-1)$.
120
121The standard deviation of the walk is then: $\mu/\sqrt(k)$.
122
123\subsection{Member Function Documentation}
124\hypertarget{classbdm_1_1mpdf_e4848a428d8ef0549c6e4a9ed386d9f2}{
125\index{bdm::migamma@{bdm::migamma}!samplecond@{samplecond}}
126\index{samplecond@{samplecond}!bdm::migamma@{bdm::migamma}}
127\subsubsection[samplecond]{\setlength{\rightskip}{0pt plus 5cm}virtual vec bdm::mpdf::samplecond (const vec \& {\em cond}, \/  double \& {\em ll})\hspace{0.3cm}{\tt  \mbox{[}inline, virtual, inherited\mbox{]}}}}
128\label{classbdm_1_1mpdf_e4848a428d8ef0549c6e4a9ed386d9f2}
129
130
131Returns a sample from the density conditioned on {\tt cond}, $x \sim epdf(rv|cond)$.
132
133\begin{Desc}
134\item[Parameters:]
135\begin{description}
136\item[{\em cond}]is numeric value of {\tt rv} \item[{\em ll}]is a return value of log-likelihood of the sample. \end{description}
137\end{Desc}
138
139
140Reimplemented in \hyperlink{classbdm_1_1mprod_1a37c2aaba8bde7fce5351c39b6e1168}{bdm::mprod}.
141
142References bdm::mpdf::condition(), bdm::mpdf::ep, bdm::epdf::evallog(), and bdm::epdf::sample().
143
144Referenced by bdm::MPF$<$ BM\_\-T $>$::bayes(), bdm::PF::bayes(), and bdm::ArxDS::step().\hypertarget{classbdm_1_1mpdf_ee26963a637b2ea1fb1933652981e652}{
145\index{bdm::migamma@{bdm::migamma}!samplecond\_\-m@{samplecond\_\-m}}
146\index{samplecond\_\-m@{samplecond\_\-m}!bdm::migamma@{bdm::migamma}}
147\subsubsection[samplecond\_\-m]{\setlength{\rightskip}{0pt plus 5cm}virtual mat bdm::mpdf::samplecond\_\-m (const vec \& {\em cond}, \/  vec \& {\em ll}, \/  int {\em N})\hspace{0.3cm}{\tt  \mbox{[}inline, virtual, inherited\mbox{]}}}}
148\label{classbdm_1_1mpdf_ee26963a637b2ea1fb1933652981e652}
149
150
151Returns.
152
153\begin{Desc}
154\item[Parameters:]
155\begin{description}
156\item[{\em N}]samples from the density conditioned on {\tt cond}, $x \sim epdf(rv|cond)$. \item[{\em cond}]is numeric value of {\tt rv} \item[{\em ll}]is a return value of log-likelihood of the sample. \end{description}
157\end{Desc}
158
159
160References bdm::mpdf::condition(), bdm::RV::count(), bdm::mpdf::ep, bdm::epdf::evallog(), bdm::mpdf::rv, and bdm::epdf::sample().
161
162The documentation for this class was generated from the following file:\begin{CompactItemize}
163\item 
164\hyperlink{libEF_8h}{libEF.h}\end{CompactItemize}
Note: See TracBrowser for help on using the browser.