Thank you Hans.
The remaining problem is that the bigger shape is scaled down to the size of the smaller shapes.


Sytse

Here is the source:

\setupFLOWcharts[
[ nx=1, ny=1,
 dx=40pt, dy=25pt,
 breedte=260pt, hoogte=130pt
]

\setupFLOWlines
[hoek=rond,
 kaderkleur=blauw,
 pijl=ja]

\startFLOWchart[Observations] % The bigger shape

\startFLOWcell
 \name{Obs}
 \location{1,1}
 \shape{action}
 \text{{\bf observations}}
\stopFLOWcell
\stopFLOWchart


\setupFLOWcharts [ nx=3, ny=2, dx=40pt, dy=25pt, breedte=110pt, hoogte=130pt ]

\startFLOWchart[Parameters] % The smaller shapes

 \startFLOWcell
  \name{Obs1}
  \location{1,1}
  \shape{node}
  \connection[bt]{CanPar}
 \stopFLOWcell

 \startFLOWcell
  \name{Obs2}
  \location{2,1}
  \shape{node}
  \connection[bt]{Means}
 \stopFLOWcell

\startFLOWcell
\name {CanPar}
\location{1,2}
\shape{action}
\text{{\bf canonical\\parameters}\\$\btheta=\pmatrix{\theta_1\cr\vdots\cr\theta_n\cr}$}
\connection[-r-l]{Means}
\comment[t]{$\mu_i=h^{-1}(\theta_i)$}
\comment[b]{$=b'(\theta_i)$}
\stopFLOWcell


\startFLOWcell
  \name {Means}
  \location{2,2}
  \shape{action}
  \text{{\bf means}\\~\\$\E\Y=\bmu=\pmatrix{\mu_1\cr\vdots\cr\mu_n\cr}$}
  \connection[+l+r]{CanPar}
  \comment[b]{$\theta_i=h(\mu_i)$}
  \connection[+r+l]{LinStruct}
  \comment[t]{link function}
  \comment[b]{$\eta_i=g(\mu_i)$}
 \stopFLOWcell

\startFLOWcell
\name {LinStruct}
\location{3,2}
\shape{action}
\text{{\bf linear structure}\\~\\$\pmatrix{\eta_i\cr\vdots\cr\eta_n\cr}={\boldmath{\eta}}=\X\bbeta$}
\connection[-l-r]{Means}
\comment[t]{$\mu_i=g^{-1}(\eta_i)$}
\comment[b]{response f.}
\stopFLOWcell


\stopFLOWchart

\startoverlay
   {\FLOWchart[Parameters]}
   {\FLOWchart[Observations]}
\stopoverlay



At 4/22/2004 11:02, you wrote:
\startoverlay
   {... de chart ...}
   {... de chart ...}
\stopoverlay


_______________________________________________
ntg-context mailing list
[EMAIL PROTECTED]
http://www.ntg.nl/mailman/listinfo/ntg-context

Reply via email to