Skip to content

Commit

Permalink
Cleanup equation notation
Browse files Browse the repository at this point in the history
  • Loading branch information
John K. Parejko committed Aug 24, 2017
1 parent 56c624f commit 1054173
Show file tree
Hide file tree
Showing 2 changed files with 16 additions and 12 deletions.
12 changes: 8 additions & 4 deletions doc/jointcal.lyx
Original file line number Diff line number Diff line change
Expand Up @@ -1939,8 +1939,12 @@ As an illustrative example, we will work through a particular photometry
\begin_inset Formula $f_{0}$
\end_inset

: the CCD filter response) and a 3th order 2-D Chebyshev polynomial (
\begin_inset Formula $\sum a_{j,k}T_{j,k}(u,v)$
: the CCD filter response) and an
\begin_inset Formula $(n+m)$
\end_inset

th order 2-D Chebyshev polynomial (
\begin_inset Formula $\sum a_{j,k}T_{j}(u)T_{k}(v)$
\end_inset

: the optics+sky response, where
Expand All @@ -1955,7 +1959,7 @@ As an illustrative example, we will work through a particular photometry
Thus, the mapping will be
\begin_inset Formula
\[
M_{\gamma i}(\eta,S_{\gamma i})=M_{CCD}(f_{0}^{-1},f_{\gamma i})M_{visit}(a_{j,k},x_{\gamma i},y_{\gamma i})=f_{\gamma i}[f_{0}]^{-1}\sum_{j=0}^{j=3}\sum_{k=0}^{k=3}a_{j,k}T_{j,k}(u_{\gamma i},v_{\gamma i})=\phi_{\gamma i}
M_{\gamma i}(\eta,S_{\gamma i})=M_{CCD}(f_{0}^{-1},f_{\gamma i})M_{visit}(a_{j,k},x_{\gamma i},y_{\gamma i})=f_{\gamma i}[f_{0}]^{-1}\sum_{j=0}^{j=n}\sum_{k=0}^{k=m}a_{j,k}T_{j}(u_{\gamma i})T_{k}(v_{\gamma i})=\phi_{\gamma i}
\]

\end_inset
Expand All @@ -1976,7 +1980,7 @@ where we will fit
Computing those derivatives gives us:
\begin_inset Formula
\begin{eqnarray*}
\nabla D_{\gamma i} & = & (\frac{\partial D_{\gamma i}}{\partial f_{0}^{-1}},\frac{\partial D_{\gamma i}}{a_{0,0}},\ldots,\frac{\partial D_{\gamma i}}{a_{3,3}},\frac{\partial D_{\gamma i}}{F_{i}})
\nabla D_{\gamma i} & = & (\frac{\partial D_{\gamma i}}{\partial f_{0}^{-1}},\frac{\partial D_{\gamma i}}{\partial a_{0,0}},\ldots,\frac{\partial D_{\gamma i}}{\partial a_{n,m}},\frac{\partial D_{\gamma i}}{\partial F_{i}})
\end{eqnarray*}

\end_inset
Expand Down
16 changes: 8 additions & 8 deletions doc/jointcal.tex
Original file line number Diff line number Diff line change
Expand Up @@ -287,9 +287,9 @@ \subsection{Choice of projectors}
In the least squares expression \ref{eq:astrometry-chi2}, the residuals
of the measurement terms read:
\[
D_{\gamma i}=M_{\gamma}(S_{\gamma,i})-P_{\gamma}(F_{j})
D_{\gamma i}=M_{\gamma}(S_{\gamma i})-P_{\gamma}(F_{i})
\]
If the coordinates $F_{j}$ are sidereal coordinates, the projector
If the coordinates $F_{i}$ are sidereal coordinates, the projector
$P_{\gamma}$ determine the meaning of the mapping $M_{\gamma}$.
If one is aiming at producing WCS's for the image, it seems wise to
choose for $P_{\gamma}$ the projection used foe the envisioned WCS,
Expand Down Expand Up @@ -492,18 +492,18 @@ \subsection{Photometry example}

As an illustrative example, we will work through a particular photometry
mapping, consisting of a constant zero-point per CCD ($f_{0}$: the
CCD filter response) and a 3th order 2-D Chebyshev polynomial ($\sum a_{j,k}T_{j,k}(u,v)$:
the optics+sky response, where $(u(x,y),v(x,y))$ are the focal plane
coordinates of pixel $(x,y)$ on a given CCD) per visit. Thus, the
mapping will be
CCD filter response) and an $(n+m)$th order 2-D Chebyshev polynomial
($\sum a_{j,k}T_{j}(u)T_{k}(v)$: the optics+sky response, where $(u(x,y),v(x,y))$
are the focal plane coordinates of pixel $(x,y)$ on a given CCD)
per visit. Thus, the mapping will be
\[
M_{\gamma i}(\eta,S_{\gamma i})=M_{CCD}(f_{0}^{-1},f_{\gamma i})M_{visit}(a_{j,k},x_{\gamma i},y_{\gamma i})=f_{\gamma i}[f_{0}]^{-1}\sum_{j=0}^{j=3}\sum_{k=0}^{k=3}a_{j,k}T_{j,k}(u_{\gamma i},v_{\gamma i})=\phi_{\gamma i}
M_{\gamma i}(\eta,S_{\gamma i})=M_{CCD}(f_{0}^{-1},f_{\gamma i})M_{visit}(a_{j,k},x_{\gamma i},y_{\gamma i})=f_{\gamma i}[f_{0}]^{-1}\sum_{j=0}^{j=n}\sum_{k=0}^{k=m}a_{j,k}T_{j}(u_{\gamma i})T_{k}(v_{\gamma i})=\phi_{\gamma i}
\]
where we will fit $f_{0}^{-1}$ instead of $f_{0}$ in order to simplify
the derivatives (with respect to $\eta=\left(f_{0}^{-1},a_{j,k}\forall j,k\right)$).
Computing those derivatives gives us:
\begin{eqnarray*}
\nabla D_{\gamma i} & = & (\frac{\partial D_{\gamma i}}{\partial f_{0}^{-1}},\frac{\partial D_{\gamma i}}{a_{0,0}},\ldots,\frac{\partial D_{\gamma i}}{a_{3,3}},\frac{\partial D_{\gamma i}}{F_{i}})
\nabla D_{\gamma i} & = & (\frac{\partial D_{\gamma i}}{\partial f_{0}^{-1}},\frac{\partial D_{\gamma i}}{\partial a_{0,0}},\ldots,\frac{\partial D_{\gamma i}}{\partial a_{n,m}},\frac{\partial D_{\gamma i}}{\partial F_{i}})
\end{eqnarray*}
where, for the measurement terms we have (recall eq. \ref{eq:photometry_residual_vector_measurement}),
\begin{eqnarray*}
Expand Down

0 comments on commit 1054173

Please sign in to comment.