immediately before and immediately after splitting are stored
to produce later output.
-\section{Analyzing output}
-
-The analysis of output is handled by a separate perl program which
-shares many common modules with the simulation program. Current output
-includes simulation progress, summary tables, summary statistics, and
-various graphs.
-
-\subsection{PCA plots}
-
-Two major groups of axes that contribute to vesicle variation between
-subsequent generations are the components and properties of vesicles.
-Each component in a vesicle is an axis on its own; it can be measured
-either as an absolute number of molecules in each component, or the
-fraction of molecules of that component over the total number of
-molecules; the second approach is often more convenient, as it allows
-vesicles with different number of molecules to be directly compared
-(though it hides any effect of vesicle size).
-
-In order to visualize the variation between and transition of
-subsequent generations of vesicles from their initial state in the
-simulation, to their final state at the termination of the simulation,
-we plot the projection of the generations onto the two principle PCA
-axes (and alternatively, any pairing of the axes).
-
-\subsection{Carpet plots}
-
-Carpet plots show the distance/similarity between the composition or
-properties of all generations in a single run. The difference between
-each group of vesicle can be calculated using Euclidean distance
-(properties and compositions) or H similarity (composition only). We
-must use Euclidean distance for properties because the H distance
-metric is invalid when comparing negative vector elements to positive
-vector elements.
-
-In addition to showing distance or similarity, carpet plots also
-depict propersomes and composomes as square boxes on the diagonals and
-propertypes and compotypes as rectangles off the diagonals, each
-propertype or compotype with a distinct color.
-
-\subsection{Propersomes, propertypes, composomes and compotypes}
-
-A generation is considered to be a propersome if it is less than $0.6$
-units (by Euclidean distance of normalized properties) away from the
-generation immediately following and preceding. Likewise, a generation
-is in a composome if its H similarity is more than $0.9$ (by the
-normalized H metric) from the preceding and following generations.
-Propersomes and composomes are then assigned to propertypes and
-compotypes using Paritioning Around Medioids
-(PAM). All values of $k$ between 2 and 15
-(or the number of propersomes and composomes, if that is less than 15)
-are tried, and the clustering with the smallest
-silhouette~\citep{Rousseeuw1987:silhouettes} is chosen as the ideal
-clustering~\citep{Shenhav2005:pgard}.
+% \section{Analyzing output}
+%
+% The analysis of output is handled by a separate perl program which
+% shares many common modules with the simulation program. Current output
+% includes simulation progress, summary tables, summary statistics, and
+% various graphs.
+%
+% \subsection{PCA plots}
+%
+% Two major groups of axes that contribute to vesicle variation between
+% subsequent generations are the components and properties of vesicles.
+% Each component in a vesicle is an axis on its own; it can be measured
+% either as an absolute number of molecules in each component, or the
+% fraction of molecules of that component over the total number of
+% molecules; the second approach is often more convenient, as it allows
+% vesicles with different number of molecules to be directly compared
+% (though it hides any effect of vesicle size).
+%
+% In order to visualize the variation between and transition of
+% subsequent generations of vesicles from their initial state in the
+% simulation, to their final state at the termination of the simulation,
+% we plot the projection of the generations onto the two principle PCA
+% axes (and alternatively, any pairing of the axes).
+%
+% \subsection{Carpet plots}
+%
+% Carpet plots show the distance/similarity between the composition or
+% properties of all generations in a single run. The difference between
+% each group of vesicle can be calculated using Euclidean distance
+% (properties and compositions) or H similarity (composition only). We
+% must use Euclidean distance for properties because the H distance
+% metric is invalid when comparing negative vector elements to positive
+% vector elements.
+%
+% In addition to showing distance or similarity, carpet plots also
+% depict propersomes and composomes as square boxes on the diagonals and
+% propertypes and compotypes as rectangles off the diagonals, each
+% propertype or compotype with a distinct color.
+%
+% \subsection{Propersomes, propertypes, composomes and compotypes}
+%
+% A generation is considered to be a propersome if it is less than $0.6$
+% units (by Euclidean distance of normalized properties) away from the
+% generation immediately following and preceding. Likewise, a generation
+% is in a composome if its H similarity is more than $0.9$ (by the
+% normalized H metric) from the preceding and following generations.
+% Propersomes and composomes are then assigned to propertypes and
+% compotypes using Paritioning Around Medioids
+% (PAM). All values of $k$ between 2 and 15
+% (or the number of propersomes and composomes, if that is less than 15)
+% are tried, and the clustering with the smallest
+% silhouette~\citep{Rousseeuw1987:silhouettes} is chosen as the ideal
+% clustering~\citep{Shenhav2005:pgard}.
\bibliographystyle{unsrtnat}
projects / ool / lipid_simulation_formalism.git / blobdiff