Restructure Interface PostProcessors

doc/user_manual/PostProcessors/HS2PS.tex

@@ -0,0 +1,17 @@
+\subsubsection{HS2PS}
+\label{HS2PS}
+
+This Post-Processor performs a conversion from HistorySet to PointSet.
+The conversion is made so that each history $H$ is converted to a single point $P$.
+Assume that each history $H$ is a dictionary (mapping) of $n$ output variables $x_1=[...],x_n=[...]$, then the resulting point $P$ is $P=concat(x_1,...,x_n)$.
+Note: it is here assumed that all histories have been synced so that they have the same length, start point and end point. 
+If you are not sure, do a pre-processing the the original history set.
+
+\ppType{HS2PS}{HS2PS}
+
+In the \xmlNode{PostProcessor} input block, the following XML sub-nodes are required,
+independent of the \xmlAttr{subType} specified (min, max, avg and value case):
+
+\begin{itemize}
+   \item \xmlNode{pivotParameter}, \xmlDesc{string, optional field}, ID of the temporal variable (only for avg)
+\end{itemize}

doc/user_manual/PostProcessors/HStoPSOperator.tex

@@ -0,0 +1,103 @@
+\subsubsection{HStoPSOperator}
+\label{HStoPSOperator}
+
+This Post-Processor performs the conversion from HistorySet to PointSet performing a projection of the output space.
+\ppType{HStoPSOperator}{HStoPSOperator}
+In the \xmlNode{PostProcessor} input block, the following XML sub-nodes are available:
+
+\begin{itemize}
+   \item \xmlNode{pivotParameter}, \xmlDesc{string, optional field}, ID of the temporal variable. Default is ``time''.
+   \nb Used just in case the  \xmlNode{pivotValue}-based operation  is requested
+    \item \xmlNode{operator}, \xmlDesc{string, optional field}, the operation to perform on the output space:
+      \begin{itemize}
+        \item \textbf{min}, compute the minimum of each variable along each single history
+         \item \textbf{max}, compute the maximum of each variable along each single history
+         \item \textbf{average}, compute the average of each variable along each single history
+         \item \textbf{all}, join together all of the each variable in
+           the history, and make the pivotParameter a regular
+           parameter.  Unlike the min and max operators, this keeps
+           all the data, just organized differently. This operator
+           does this by propagating the other input parameters for
+           each item of the pivotParameter.
+           Table~\ref{operator_all_switch_before} shows an example
+           HistorySet with input parameter x, pivot parameter t, and
+           output parameter b and then
+           Table~\ref{operator_all_switch_after} shows the resulting
+           PointSet with input parameters x and t, and output
+           parameter b. Note that which parameters are input and which
+           are output in the resulting PointSet depends on the
+           DataObject specification.
+       \end{itemize}
+        \nb This node can be inputted only if \xmlNode{pivotValue} and \xmlNode{row} are not present
+     \item \xmlNode{pivotValue}, \xmlDesc{float, optional field}, the value of the pivotParameter with respect to the other outputs need to be extracted.
+       \nb This node can be inputted only if \xmlNode{operator} and \xmlNode{row} are not present
+     \item \xmlNode{pivotStrategy}, \xmlDesc{string, optional field}, The strategy to use for the pivotValue:
+       \begin{itemize}
+        \item \textbf{nearest}, find the value that is the nearest with respect the \xmlNode{pivotValue}
+        \item \textbf{floor}, find the value that is the nearest with respect to the \xmlNode{pivotValue} but less then the  \xmlNode{pivotValue}
+        \item \textbf{celing}, find the value that is the nearest with respect to the \xmlNode{pivotValue} but greater then the  \xmlNode{pivotValue}
+        \item \textbf{interpolate}, if the exact  \xmlNode{pivotValue}  can not be found, interpolate using a linear approach
+       \end{itemize}
+
+       \nb Valid just in case \xmlNode{pivotValue} is present
+     \item \xmlNode{row}, \xmlDesc{int, optional field}, the row index at which the outputs need to be extracted.
+       \nb This node can be inputted only if \xmlNode{operator} and \xmlNode{pivotValue} are not present
+\end{itemize}
+
+This example will show how the XML input block would look like:
+
+\begin{lstlisting}[style=XML,morekeywords={subType,debug,name,class,type}]
+<Simulation>
+  ...
+  <Models>
+    ...
+    <PostProcessor name="HStoPSperatorRows" subType="HStoPSOperator">
+      <row>-1</row>
+    </PostProcessor>
+    <PostProcessor name="HStoPSoperatorPivotValues" subType="HStoPSOperator">
+        <pivotParameter>time</pivotParameter>
+        <pivotValue>0.3</pivotValue>
+    </PostProcessor>
+    <PostProcessor name="HStoPSoperatorOperatorMax" subType="HStoPSOperator">
+        <pivotParameter>time</pivotParameter>
+        <operator>max</operator>
+    </PostProcessor>
+    <PostProcessor name="HStoPSoperatorOperatorMin" subType="HStoPSOperator">
+        <pivotParameter>time</pivotParameter>
+        <operator>min</operator>
+    </PostProcessor>
+    <PostProcessor name="HStoPSoperatorOperatorAverage" subType="HStoPSOperator">
+        <pivotParameter>time</pivotParameter>
+        <operator>average</operator>
+    </PostProcessor>
+    ...
+  </Models>
+  ...
+</Simulation>
+\end{lstlisting}
+
+\begin{table}[!hbtp]
+  \caption{Starting HistorySet for operator all}
+  \label{operator_all_switch_before}
+\begin{tabular}{l|l|l}
+  x & t & b \\
+  \hline
+  5.0 &  &  \\
+  \hline
+  & 1.0 & 6.0 \\
+  \hline
+  & 2.0 & 7.0 \\
+\end{tabular}
+\end{table}
+
+\begin{table}[!hbtp]
+  \caption{Resulting PointSet after operator all}
+  \label{operator_all_switch_after}
+\begin{tabular}{l|l|l}
+  x & t & b \\
+  \hline
+  5.0 & 1.0 & 6.0  \\
+  \hline
+  5.0 & 2.0 & 7.0 \\
+\end{tabular}
+\end{table}

doc/user_manual/PostProcessors/HistorySetSampling.tex

@@ -0,0 +1,35 @@
+\subsubsection{HistorySetSampling}
+\label{HistorySetSampling}
+
+This Post-Processor performs the conversion from HistorySet to HistorySet
+The conversion is made so that each history H is re-sampled accordingly  to a
+specific sampling strategy.
+It can be used to reduce the amount of space required by the HistorySet.
+
+\ppType{HistorySetSampling}{HistorySetSampling}
+
+In the \xmlNode{PostProcessor} input block, the following XML sub-nodes are required,
+independent of the \xmlAttr{subType} specified:
+
+\begin{itemize}
+   \item \xmlNode{samplingType}, \xmlDesc{string, required field}, specifies the type of sampling method to be used:
+   \begin{itemize}
+     \item uniform: the set of \xmlNode{numberOfSamples} samples are uniformly distributed along the time axis
+     \item firstDerivative: the set of \xmlNode{numberOfSamples} samples are distributed along the time axis in regions with
+                            higher first order derivative
+     \item secondDerivative: the set of \xmlNode{numberOfSamples} samples are distributed along the time axis in regions with
+                             higher second order derivative
+     \item filteredFirstDerivative: samples are located where the first derivative is greater than the specified \xmlNode{tolerance} value
+                                    (hence, the number of samples can vary from history to history)
+     \item filteredSecondDerivative: samples are located where the second derivative is greater than the specified \xmlNode{tolerance} value
+                                     (hence, the number of samples can vary from history to history)
+   \end{itemize}
+   \item \xmlNode{numberOfSamples}, \xmlDesc{integer, optional field}, number of samples (required only for the following sampling
+                                             types: uniform, firstDerivative secondDerivative)
+   \item \xmlNode{pivotParameter}, \xmlDesc{string, required field}, ID of the temporal variable
+   \item \xmlNode{interpolation}, \xmlDesc{string, optional field}, type of interpolation to be employed for the history reconstruction
+                                           (required only for the following sampling types: uniform, firstDerivative secondDerivative).
+                                           Valid types of interpolation to specified: linear, nearest, zero, slinear, quadratic, cubic, intervalAverage
+   \item \xmlNode{tolerance}, \xmlDesc{string, optional field}, tolerance level (required only for the following sampling types:
+                                       filteredFirstDerivative or filteredSecondDerivative)
+\end{itemize}

doc/user_manual/PostProcessors/HistorySetSnapShot.tex

@@ -0,0 +1,94 @@
+\subsubsection{HistorySetSnapShot}
+\label{HistorySetSnapShot}
+
+This Post-Processor performs a conversion from HistorySet to PointSet.
+The conversion is made so that each history $H$ is converted to a single point $P$.
+There are several methods that can be employed to choose the single point from the history:
+\begin{itemize}
+  \item min: Take a time slice when the \xmlNode{pivotVar} is at its smallest value,
+  \item max: Take a time slice when the \xmlNode{pivotVar} is at its largest value,
+  \item average: Take a time slice when the \xmlNode{pivotVar} is at its time-weighted average value,
+  \item value: Take a time slice when the \xmlNode{pivotVar} \emph{first passes} its specified value,
+  \item timeSlice: Take a time slice index from the sampled time instance space.
+\end{itemize}
+To demonstrate the timeSlice, assume that each history H is a dict of n output variables $x_1=[...],
+x_n=[...]$, then the resulting point P is at time instant index t: $P=[x_1[t],...,x_n[t]]$.
+
+Choosing one the these methods for the \xmlNode{type} node will take a time slice for all the variables in the
+output space based on the provided parameters.  Alternatively, a \xmlString{mixed} type can be used, in which
+each output variable can use a different time slice parameter.  In other words, you can take the max of one
+variable while taking the minimum of another, etc.
+
+\ppType{HistorySetSnapShot}{HistorySetSnapShot}
+
+In the \xmlNode{PostProcessor} input block, the following XML sub-nodes are required,
+independent of the \xmlAttr{subType} specified:
+
+\begin{itemize}
+  \item \xmlNode{type}, \xmlDesc{string, required field}, type of operation: \xmlString{min}, \xmlString{max},
+                        \xmlString{average}, \xmlString{value}, \xmlString{timeSlice}, or \xmlString{mixed}
+   \item \xmlNode{extension}, \xmlDesc{string, required field}, type of extension when the sync process goes outside the boundaries of the history (zeroed or extended)
+   \item \xmlNode{pivotParameter}, \xmlDesc{string, optional field}, name of the temporal variable.  Required for the
+     \xmlString{average} and \xmlString{timeSlice} methods.
+\end{itemize}
+
+If a \xmlString{timeSlice} type is in use, the following nodes also are required:
+\begin{itemize}
+   \item \xmlNode{timeInstant}, \xmlDesc{integer, required field}, required and only used in the
+     \xmlString{timeSlice} type.  Location of the time slice (integer index)
+   \item \xmlNode{numberOfSamples}, \xmlDesc{integer, required field}, number of samples
+\end{itemize}
+
+If instead a \xmlString{min}, \xmlString{max}, \xmlString{average}, or \xmlString{value} is used, the following nodes
+are also required:
+\begin{itemize}
+   \item \xmlNode{pivotVar}, \xmlDesc{string, required field},  Name of the chosen indexing variable (the
+         variable whose min, max, average, or value is used to determine the time slice)
+       \item \xmlNode{pivotVal}, \xmlDesc{float, optional field},  required for \xmlString{value} type, the value for the chosen variable
+\end{itemize}
+
+Lastly, if a \xmlString{mixed} approach is used, the following nodes apply:
+\begin{itemize}
+  \item \xmlNode{max}, \xmlDesc{string, optional field}, the names of variables whose output should be their
+    own maximum value within the history.
+  \item \xmlNode{min}, \xmlDesc{string, optional field}, the names of variables whose output should be their
+    own minimum value within the history.
+  \item \xmlNode{average}, \xmlDesc{string, optional field}, the names of variables whose output should be their
+    own average value within the history. Note that a \xmlNode{pivotParameter} node is required to perform averages.
+  \item \xmlNode{value}, \xmlDesc{string, optional field}, the names of variables whose output should be taken
+    at a time slice determined by another variable.  As with the non-mixed \xmlString{value} type, the first
+    time the \xmlAttr{pivotVar} crosses the specified \xmlAttr{pivotVal} will be the time slice taken.
+    This node requires two attributes, if used:
+    \begin{itemize}
+      \item \xmlAttr{pivotVar}, \xmlDesc{string, required field}, the name of the variable on which the time
+        slice will be performed.  That is, if we want the value of $y$ when $t=0.245$,
+        this attribute would be \xmlString{t}.
+      \item \xmlAttr{pivotVal}, \xmlDesc{float, required field}, the value of the \xmlAttr{pivotVar} on which the time
+        slice will be performed.  That is, if we want the value of $y$ when $t=0.245$,
+        this attribute would be \xmlString{0.245}.
+    \end{itemize}
+  Note that all the outputs of the \xmlNode{DataObject} output of this postprocessor must be listed under one
+  of the \xmlString{mixed} node types in order for values to be returned.
+\end{itemize}
+
+\textbf{Example (mixed):}
+This example will output the average value of $x$ for $x$, the value of $y$ at
+time$=0.245$ for $y$, and the value of $z$ at $x=4.0$ for $z$.
+\begin{lstlisting}[style=XML,morekeywords={subType,debug,name,class,type}]
+<Simulation>
+  ...
+  <Models>
+    ...
+    <PostProcessor name="mampp2" subType="HistorySetSnapShot">
+      <type>mixed</type>
+      <average>x</average>
+      <value pivotVar="time" pivotVal="0.245">y</value>
+      <value pivotVar="x" pivotVal="4.0">z</value>
+      <pivotParameter>time</pivotParameter>
+      <extension>zeroed</extension>
+    </PostProcessor>
+    ...
+  </Models>
+  ...
+</Simulation>
+\end{lstlisting}

doc/user_manual/PostProcessors/HistorySetSync.tex

@@ -0,0 +1,32 @@
+\subsubsection{HistorySetSync}
+\label{HistorySetSync}
+
+This Post-Processor performs the conversion from HistorySet to HistorySet
+The conversion is made so that all histories are synchronized in time.
+It can be used to allow the histories to be sampled at the same time instant.
+
+There are two possible synchronization methods, specified through the \xmlNode{syncMethod} node.  If the
+\xmlNode{syncMethod} is \xmlString{grid}, a \xmlNode{numberOfSamples} node is specified,
+which yields an equally-spaced grid of time points. The output values for these points will be linearly derived
+using nearest sampled time points, and the new HistorySet will contain only the new grid points.
+
+The other methods are used by specifying \xmlNode{syncMethod} as \xmlString{all}, \xmlString{min}, or
+\xmlString{max}.  For \xmlString{all}, the postprocessor will iterate through the
+existing histories, collect all the time points used in any of them, and use these as the new grid on which to
+establish histories, retaining all the exact original values and interpolating linearly where necessary.
+In the event of \xmlString{min} or \xmlString{max}, the postprocessor will find the smallest or largest time
+history, respectively, and use those time values as nodes to interpolate between.
+
+\ppType{HistorySetSync}{HistorySetSync}
+
+In the \xmlNode{PostProcessor} input block, the following XML sub-nodes are required,
+independent of the \xmlAttr{subType} specified:
+
+\begin{itemize}
+   \item \xmlNode{pivotParameter}, \xmlDesc{string, required field}, ID of the temporal variable
+   \item \xmlNode{extension}, \xmlDesc{string, required field}, type of extension when the sync process goes outside the boundaries of the history (zeroed or extended)
+   \item \xmlNode{syncMethod}, \xmlDesc{string, required field}, synchronization strategy to employ (see
+     description above).  Options are \xmlString{grid}, \xmlString{all}, \xmlString{max}, \xmlString{min}.
+   \item \xmlNode{numberOfSamples}, \xmlDesc{integer, optional field}, required if \xmlNode{syncMethod} is
+     \xmlString{grid}, number of new time samples
+\end{itemize}