Boleslaw Szymanski and Thomas Caraco. Spatial analysis of vector-borne disease: A four species model. Evolutionary Ecology, 8(3):299-314, 1994.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....components with out ports. For a simulation to be linkable, it should have outports that are used by the simulation to produce data available to other components and or inports that 27 require additional information (often dynamic) during execution. For example, a Lyme disease simulation [24, 64] can be constructed as a composition of two com ponents, one simulating mice and the other simulating ticks. Ticks are immobile but they can be carried around and dropped by mice. The Mice component there fore needs to access the tick density at the location where an event related to mice ....

Boleslaw Szymanski and Thomas Caraco. Spatial analysis of vector-borne disease: A four species model. Evolutionary Ecology, 8(3):299-314, 1994.

....: 8 2 List of Tables 3 Chapter 1 Introduction The study of spatial and temporal aspects of multi species biological systems is central to ecology. Epidemics have special significance in terms of agriculture, public health and ecology. Szymanski and Caraco [4] have developed a cellular automata for spatially explicit modeling of population dynamics and epidemics. A software package named TEMPEST implements this model on several high performance (parallel) computers [1, 2, 3] additional information is available on line. This document describes the use ....

B. K. Szymanski and T. Caraco. Spatial analysis of vector-borne disease: A four species model. Evolutionary Ecology, 8:299--314, 1994. Online information available at ftp://ftp.cs.rpi.edu/pub/szymansk/eec.ps. 14

....Implementation of TEMPEST Version 3. 0 Bill Maniatty y Department of Computer Science Troy, NY 12180 3590 1 Introduction This document describes the user interface of the implementation of TEMPEST for spatially explicit modeling of ecological phenomena as detailed in Szymanski and Caraco, [3]. The structure of the document is as follows: ffl How to Install the Software on a MasPar MP 1 or MP 2 in Section 2 ffl How to Invoke the Software in Section 3 ffl Parameter file format in Section 5 ffl Configuration file format in Section 6 ffl Interval file format in Section 7 ffl Site ....

....and comments. Comments are started by a # symbol, and extend to the end of line. 5 Parameter File Format The parameter file is human readable, and is parsed at run time. The parameters contained in the file specify: ffl the weights used in computing state transition probabilities, as found in [3]. ffl sequencing of transitions within the discrete time simulation tick. ffl version dependent information, we reserve the right in the future to allow additional controls to be added to the simulation, and to maintain backward compatibility we employ a version number at the start of the ....

[Article contains additional citation context not shown here]

B. K. Szymanski and T. Caraco. Spatial analysis of vector-borne disease: A four species model. Evolutionary Ecology, 8:299--314, 1994.

....placement of data and operations in large scientific computations executed on a distributed memory machine can significantly reduce the total computation time. In this paper we consider the impact of operation placement on the efficiency of computations that arise in theoretical population biology [1]. The computation that motivated the study of simultaneous reduction algorithms models an isolated ecosystem by simulating the dynamics of interspecies interactions. The model represents the ecosystem as a cellular automata whose cells correspond to the habitat sites that are small enough to ....

.... differential equations characterizing fluid flow) Parallel Reduction is defined as the application of an associative binary operator across a list of elements [3] Given an array, A, of size n and the binary operator, parallel reduction obtains the scalar result, A[i] as follows: kA[i] A[1] A[2] A[3] A[n] Simultaneous Parallel Reduction (SPR) applies parallel reduction concurrently over many overlapping sections of an array. Given an array A[1. N] a simultaneous parallel reduction for array sections of size n with an offset of the result r, 0 r n, obtains an array ....

[Article contains additional citation context not shown here]

B. Szymanski and T. Caraco, Spatial analysis of vector-borne disease: a four species model, Evolutionary Ecology, 8 (1994) 299-314.

....of results can only be found in spatially explicit models and cannot be generated by spatially homogeneous models. 3 Implementation and Performance of STORM The STORM model was implemented using C and MPI much like the TEM PEST model described in [14] Prior performance analysis of TEMPEST [17, 14] was done on a SIMD mesh architecture (a MasPar MP 1) a network of worksta lambda = 0.2, sigma = 0.01: Time vs. SO Infection Density 0.1 3x3 SO 0 200 400 600 800 100012001400160018002000 (a) Strain with low virulence, 0.01 lambda = 0.2, sigma = 0.01 : Time vs. S1 Infection Density (b) ....

....is accomplished using the genetic algorithms on the genetic infor mation, copying, mutation, and (where applicable) crossover. Hosts are modeled individually and sites are small enough to contain at most one host, in the same fashion as the model presented here and the Szymanski Caraco model [17]. The host species are assumed to have sexual reproduction (and hence crossover of genetic material) while the parasitic species are treated as asexual. Such a model poses novel computational challenges, including: computing the lineage of each organism, measuring the simulation trajectory given ....

B. K. Szymanski and T. Caraco. Spatial analysis of vector-borne disease: A four species model. Evolutionary Ecology, 8:299-314, 1994. Online information available at ft p://ftp.cs.rpi.edu/pub/szymansk/eec.ps.

....we are beginning our investigations by considering a single genotype with multiple phenotypes, and in particular we consider the simplest form, analyzing extinction trends in a system governed by the interaction between two phenotypes. This is similar to our approaches to modeling selection in [6, 13], in the future we plan to extend our approach to model mutation, and selection much like in [8] We consider two different models of selection in RNA worlds, first beginning with a two dimensional spatially explicit individual based model, and then exploring a spatially explicit aggregation based ....

....individual based model, and then exploring a spatially explicit aggregation based model. 3. 1 An Individual Based Modeling Approach The individual based model partitions a two dimensional space into a lattice of J 1 sites and synchronously advances time using uniformly sized time steps, like [6, 13]. It is assumed that RNA replicator molecules in this model have the same sequence (i.e. are of a common genotype G) and can belong to one of two phenotypes. Sites in this model are sufficiently small that they can hold at most one replicator. Uninhabited sites are partitioned into two groups, ....

B. K. Szymanski and T. Caraco. Spatial analysis of vector-borne disease: A four species model. Evolutionary Ecology, 8:299-314, 1994.

....interaction implemented on a massively parallel computer. Many natural processes, including competition, can be characterized by a global state that is defined as a vector of local states. Correspondingly, the composition of local state transitions determine the global state tran sition [Szymanski and Caraco 1993]. To simulate the spatial and temporal dynamics of a multi species interaction, their environment is partitioned into a grid of sites; each site can support at most one organism. Each site s state indicates the presence or absence of species involved in a simulation. At each time step our ....

....Consequently, our simulations can address the question of how behavioral selectivity influences the dynamics of an epidemic and so governs spatial pattern in an interactive biotic community. The formal definition of the probabilistic model on which the implementation is based is presented in [Szymanski and Caraco 1993]. Here, the discussion fo cuses on those features of the model that affect the struc ture of the implementation. The model partitions a two dimensional environment into sites. Each site can support at most one host organism. There are two types of hosts: the ecologically dominant species is the ....

[Article contains additional citation context not shown here]

Szymanski, B., & T. Caraco, "Spatial Analysis of Vector-Borne Disease: A Four Species Model," in press for Evolutionary Ecology.

....reflects the strain s capacity to exploit its host, and so controls the rate at which it converts host re sources to its own reproduction. A model that we proposed here is an extension of the spattally explicit individual based cellular automata model that we developed for vector born dis eases [18]. The model enables the user to define an interaction neighborhoods for each site. This neighborhood, termed the ecological stencil, may extended beyond the nearest neighbors of the site. When the area of the ecological stencil approaches the environment size, the model tends to behave in a ....

....hosts resources and can dampen host reproductive intensity. Likewise infected hosts may reproduce at a reduced rate in response to their infected condition. 2. 10 Site Propagation Probability Computations Given the individual rate of propagule offspring creation at each site, the formulas given in [18] can be modified to take into account the reproduction (now heterogeneous) To account for the interactions be tween the competing species, we consider also the post dispersal competition [18] 2.11 Determining a host offspring s genotype The first step in computing a host offspring s genotype ....

[Article contains additional citation context not shown here]

B. K. Szymanski and T. Caraco. Spatial analy- sis of vector-borne disease: A four species model. Evolutionary Ecology, 8:299-314, 1994.

....ecology AMS subject classifications. 68Q22, 68U05, 92D25, 92D30 1. Introduction. Many large scale natural phenomena arise from concurrent, spatially local ized interactions of small entities. Examples are fluid dynamics [46] forest fires [14] population dynamics and spread of epidemics [44]. Cellular automata (CA) model such phenomena by discretizing space into a lattice of cells [9] Each cell is in one of a finite set of discrete states and interacts with its neighborhood, referred to as the cell s stencil. In CA models each cell updates its value as a function of its current ....

.... important to understanding population dynamics, community stability, and biodiversity [23, 25] To simulate the spatial and temporal dynamics of a multi species ecosystem, the habitat is mapped onto a GCA with each cell s state indicating the presence or absence of species involved in a simulation [44]. A probabilistic model can describe a cell s state transitions as functions of the states of cells within its stencil. The size of the stencil and the location of the affected cell within it are based on the characteristics of the species and the habitat. The implemented model allows di verse ....

[Article contains additional citation context not shown here]

B. K. SZYMANSKI AND T. CARACO, Spatial analysis of vector-borne disease: A four species model, Evolutionary Ecology, 8 (1994), pp. 299-314. Online information available at ftp://ftp.cs.rpi.edu/pub/szymansk/eec.ps.

....total execution time. The above computation is an example of a reduction evaluated simultaneously over many overlapping continuous sections of an array. Other examples of usage of such operations are likely to be found in cluster recognition, fractal dimension computation in biological modeling [34], or in modeling physical phenomena (e.g. solvers of partial differential equations characterizing fluid flow) Simultaneous reduction is evaluated over a one dimensional consecutive section of an array, called here an array interval; each array element is used as an operand to many reductions ....

B. Szymanski and T. Caraco, "Spatial analysis of vector-borne disease: A four species model," Evolutionary Ecology, 1994, vol. 8. in press.

....placement of data and operations in large scientific computations executed on a distributed memory machine can significantly reduce the total computation time. In this paper we consider the impact of operation placement on the efficiency of computations that arise in theoretical population biology [1]. The computation that motivated the study of simultaneous reduction algorithms models an isolated ecosystem by simulating the dynamics of interspecies interactions. The model represents the ecosystem as a cellular automata whose cells correspond to the habitat sites that are small enough to ....

.... fluid flow) Parallel Reduction is defined as the application of an associative binary operator across a list of elements [3] Given an array, A, of size n and the binary operator, Phi, parallel reduction obtains the scalar result, Phi n i=1 A[i] as follows: Phi n i=1 A[i] A[1] Phi A[2] Phi A[3] Phi : Phi A[n] Simultaneous Parallel Reduction (SPR) applies parallel reduction concurrently over many overlapping sections of an array. Given an array A[1: N ] a simultaneous parallel reduction for array sections of size n with an offset of the result r; 0 r n, ....

[Article contains additional citation context not shown here]

B. Szymanski and T. Caraco, Spatial analysis of vector-borne disease: a four species model, Evolutionary Ecology, 8 (1994) 299--314.

....ecological interaction implemented on a massively parallel computer. Many natural processes, including competition, can be characterized by a global state that is defined as a vector of local states. Correspondingly, the composition of local state transitions determine the global state transition [Szymanski and Caraco 1993]. To simulate the spatial and temporal dynamics of a multi species inter This work was sponsored in part by ONR under Grant N0001493 J 0076 and National Science Foundation under Grant ASC9318184. The content of the information does not necessarily reflect the position or the policy of the ....

....Consequently, our simulations can address the question of how behavioral selectivity influences the dynamics of an epidemic and so governs spatial pattern in an interactive biotic community. The formal definition of the probabilistic model on which the implementation is based is presented in [Szymanski and Caraco 1993]. Here, the discussion focuses on those features of the model that affect the structure of the implementation. The model partitions a two dimensional environment into sites. Each site can support at most one host organism. There are two types of hosts: the ecologically dominant species is the ....

[Article contains additional citation context not shown here]

Szymanski, B., & T. Caraco, "Spatial Analysis of Vector-Borne Disease: A Four Species Model," in press for Evolutionary Ecology.

....of operations to be performed on the data values from the definition of data mapping and synchronization needed to supply these data values to the proper processor at the proper instance of the program execution. With this goal in mind, we developed tools (Govindaraju and Szymanski, 1992; Szymanski, 1994) that (i) decompose, at least partially, the parallel program into the two (nearly) orthogonal parts described above, ii) translate the necessary data movements into optimal form customized for the target architecture, and (iii) synthesize an overall parallel computation. Using these tools the ....

....is intended as an overview of the research done towards implementing software tools as envisioned in Figure 1. More technical discussion can be found elsewhere (Bruno and Szymanski, 1988; Ozturan et al. 1992; Spier and Szymanski, 1990; Szymanski, 1991; Szymanski and Prywes, 1988; Sinharoy and Szymanski, 1994a; Sinharoy and Szymanski, 1994b; McKenney and Szymanski, 1992) A brief description of the EPL language, its annotations and configurations is given in Section 2. The relationship of EPL constructs and tools to different levels of parallelism is discussed in Section 3. The EPL compiler is ....

[Article contains additional citation context not shown here]

Szymanski, B. and Caraco, T. (1994). Spatial analysis of vector-borne disease: A four species model. Evolutionary Ecology, 8, in press.

....placement of data and operations in large scientific computations executed on a distributed memory machine can significantly reduce the total computation time. In this paper we consider the impact of operation placement on the efficiency of computations that arise in theoretical population biology [1]. The computation that motivated the study of simultaneous reduction algorithms models an isolated ecosystem by simulating the dynamics of multispecies interactions. The model represents the ecosystem as a cellular automata whose cells correspond to the habitat sites that are small enough to ....

.... flow) Parallel Reduction is defined as the application of an associative binary operator across a list of elements [4,5,6] Given an array A of size n and the binary operator Phi, parallel reduction obtains the scalar result, r, as follows: r = reduce(A; 1; n; Phi) Phi n i=1 A[i] A[1] Phi A[2] Phi A[3] Phi : Phi A[n] Simultaneous Parallel Reduction applies parallel reduction concurrently over many overlapping sections of an array. Given an array A[1: N ] a simultaneous parallel reduction for array sections of size n with an offset of the result r; 0 r n, obtains an ....

[Article contains additional citation context not shown here]

B. Szymanski and T. Caraco, Spatial analysis of vector-borne disease: a four species model, Evolutionary Ecology, 8 (1994) 299--314.

....not necessarily reflect the position or policy of the U.S. Government no official endorsements should be inferred or implied. Many natural systems, including epidemics, can be characterized by a global state that is defined as a vector of local states. Correspondingly, in the discussed model [5], the composition of local transitions determines each global state transition. To simulate the spatial and temporal dynamics of a multi species ecosystem, the model partitions its habitat into a grid of sites, such that each site has space for at most one host. Each site s state indicates the ....

....parallel SIMD machine, are presented in Section 4. The ecological results are interpreted in Section 5 and conclusions and future directions are discussed in Section 6. 2. The Mathematical Model The formal definition of the probabilistic model on which the implementation is based is presented in [5]. Here, we briefly examine those features of the model that affect the structure of the implementation. The model partitions a two dimensional habitat into sites. Each site can support at most one host organism. There are two types of hosts: the ecologically dominant species are the good hosts, ....

Caraco, T. & Szymanski, B. 1994. Spatial Analysis of Vector-Borne Disease: A Four Species Model. Evolutionary Ecology, in press.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC