arXiv subjects
Publications and source records attributed to Paulo Murilo Castro de Oliveira.
A dynamic agent model is introduced with an annual random wealth multiplicative process followed by taxes paid according to a linear wealth-dependent tax rate. If poor agents pay higher tax rates than rich agents, eventually all wealth becomes concentrated in the hands of a single agent. By contrast, if poor agents are subject to lower tax rates, the economic collective process continues forever.
In this work, we present a comparative study of the accuracy provided by the Wang-Landau sampling and the Broad Histogram method to estimate de density of states of the two dimensional Ising ferromagnet. The microcanonical averages used to describe the thermodynamic behaviour and to use the Broad Histogram method were obtained using the single spin-flip Wang-Landau sampling, attempting to convergence issues and accuracy improvements. We compare the results provided by both techniques with the exact ones for thermodynamic properties and critical exponents. Our results, within the Wang-Landau sampling, reveal that the Broad Histogram approach provides a better description of the density of states for all cases analysed.
Flow of viscous fluids are not usually discussed in detail in general and basic courses of physics. This is due in part to the fact that the Navier-Stokes equation has analytical solution only for a few restricted cases, while more sophisticated problems can only be solved by numerical methods. In this text, we present a computer simulation of wind tunnel, i.e., we present a set of programs to solve the Navier-Stokes equation for an arbitrary object inserted in a wind tunnel. The tunnel enables us to visualize the formation of vortices behind object, the so-called von Kármán vortices, and calculate the drag force on the object. We believe that this numerical wind tunnel can support the teacher and allow a more elaborate discussion of viscous flow. The potential of the tunnel is exemplified by the study of the drag on a simplified model of wing whose angle of attack can be controlled. A link to download the programs that make up the tunnel appears at the end.
In this work we study opinion formation in a population participating of a public debate with two distinct choices. We considered three distinct mechanisms of social interactions and individuals' behavior: conformity, nonconformity and inflexibility. The conformity is ruled by the majority-rule dynamics, whereas the nonconformity is introduced in the population as an independent behavior, implying the failure to attempted group influence. Finally, the inflexible agents are introduced in the population with a given density. These individuals present a singular behavior, in a way that their stubbornness makes them reluctant to change their opinions. We consider these effects separately and all together, with the aim to analyze the critical behavior of the system. We performed numerical simulations in some lattice structures and for distinct population sizes, and our results suggest that the different formulations of the model undergo order-disorder phase transitions in the same universality class of the Ising model. Some of our results are complemented by analytical calculations.
Street demonstrations occur across the world. In Rio de Janeiro, June/July 2013, they reach beyond one million people. A wrathful reader of \textit{O Globo}, leading newspaper in the same city, published a letter \cite{OGlobo} where many social questions are stated and answered Yes or No. These million people of street demonstrations share opinion consensus about a similar set of social issues. But they did not reach this consensus within such a huge numbered meetings. Earlier, they have met in diverse small groups where some of them could be convinced to change mind by other few fellows. Suddenly, a macroscopic consensus emerges. Many other big manifestations are widespread all over the world in recent times, and are supposed to remain in the future. The interesting questions are: 1) How a binary-option opinion distributed among some population evolves in time, through local changes occurred within small-group meetings? and 2) Is there some natural selection rule acting upon? Here, we address these questions through an agent-based model.
An evolutionary tree is a cascade of bifurcations starting from a single common root, generating a growing set of daughter species as time goes by. Species here is a general denomination for biological species, spoken languages or any other entity evolving through heredity. From the N currently alive species within a clade, distances are measured through pairwise comparisons made by geneticists, linguists, etc. The larger is such a distance for a pair of species, the older is their last common ancestor. The aim is to reconstruct the past unknown bifurcations, i.e. the whole clade, from the knowledge of the N(N-1)/2 quoted distances taken for granted. A mechanical method is presented, and its applicability discussed.
In this work we study a modified version of the majority-vote model with noise. In particular, we consider a random diluted square lattice for which a site is empty with a probability $r$. In order to analyze the critical behavior of the model, we perform Monte Carlo simulations on lattices with linear sizes up to L=140. By means of a finite-size scaling analysis we estimate the critical noises $q_{c}$ and the critical ratios $β/ν$, $γ/ν$ and $1/ν$ for some values of the probability $r$.
In this work we study a modified version of the two-dimensional Sznajd sociophysics model. In particular, we consider the effects of agents' reputations in the persuasion rules. In other words, a high-reputation group with a common opinion may convince their neighbors with probability $p$, which induces an increase of the group's reputation. On the other hand, there is always a probability $q=1-p$ of the neighbors to keep their opinions, which induces a decrease of the group's reputation. These rules describe a competition between groups with high reputation and hesitant agents, which makes the full-consensus states (with all spins pointing in one direction) more difficult to be reached. As consequences, the usual phase transition does not occur for $p<p_{c} \sim 0.69$ and the system presents realistic democracy-like situations, where the majority of spins are aligned in a certain direction, for a wide range of parameters.
A small and light polystyrene ball is released without initial speed from a certain height above the floor. Then, it falls on air. The main responsible for the friction force against the movement is the wake of successive air vortices which form behind (above) the falling ball, a turbulent phenomenon. After the wake appears, the friction force compensates the Earth gravitational attraction and the ball speed stabilises in a certain limiting value Vl. Before the formation of the turbulent wake, however, the friction force is not strong enough, allowing the initially growing speed to surpass the future final value Vl. Only after the wake finally becomes long enough, the ball speed decreases and reaches the proper Vl.
This paper presents Monte Carlo simulations of language populations and the development of language families, showing how a simple model can lead to distributions similar to the ones observed empirically. The model used combines features of two models used in earlier work by phycisists for the simulation of competition among languages: the "Viviane" model for the migration of people and propagation of languages and the "Schulze" model, which uses bitstrings as a way of characterising structural features of languages.
We study the genetic behaviour of a population formed by haploid individuals which reproduce asexually. The genetic information for each individual is stored along a bit-string (or chromosome) with L bits, where 0-bits represent the wild-type allele and 1-bits correspond to harmful mutations. Each newborn inherits this chromosome from its parent with some few random mutations: on average a fixed number m of bits are flipped. Selection is implemented according to the number N of 1-bits counted along the individual's chromosome: the smaller N the higher the probability an individual has to survive a new time step. Such a population evolves, with births and deaths, and its genetic distribution becomes stabilised after many enough generations have passed. The question we pose concerns the procedure of increasing L. The aim is to get the same distribution of relative genetic loads N/L among the equilibrated population, in spite of a larger L. Should we keep the same mutation rate m/L for different values of L? The answer is yes, which intuitively seems to be plausible. However, this conclusion is not trivial, according to our simulational results: the question involves also the population size.
The purpose of this text is: 1) to clarify the foundations of the broad histogram method, stressing the conceptual differences between it and reweighting procedures in general; 2) to propose a very simple microcanonical dynamic rule, yet to be tested by theoretical grounds, which could provide a good improvement to numerical simulations.
Computer modelling for evolutionary systems consists in: 1) to store in the memory the individual features of each member of a large population; and 2) to update the whole system repeatedly, as time goes by, according to some prescribed rules (reproduction, death, ageing, etc) where some degree of randomness is included through pseudo-random number sequences. Compared to direct observation of Nature, this approach presents two distinguishing features. First, one can follow the characteristics of the system in real time, instead of only observing the current, static situation which is a long-term consequence of a remote past completely unknown except for some available fossil snapshots. In particular, one can repeat the whole dynamical process, starting from the same initial population, using the same randomness, changing only some minor contingency during the process, in order to study its long-term consequences. Second, evolution necessarily follows a critical dynamics with long-term memory characteristics, equivalent to the long-range correlations responsible for the well known universality properties of static critical phenomena. Accordingly, some strong simplifications can be applied, allowing one to obtain many characteristics of real populations from toy models easily implementable on the computer.
The long-term behaviour of dynamic systems can be classified in two different regimes, regular or chaotic, depending on the values of the control parameters, which are kept constant during the time evolution. Starting from slightly different initial conditions, a regular system converges to the same final trajectory, whereas a chaotic system follows two distinct trajectories exponentially diverging from each other. In spite of these differences, regular and chaotic systems share a common property: both arrive exponentially fast to their final destiny, becoming trapped there. In both cases one has finite transient times. This is not a profitable property in what concerns evolutionary strategies, where the eternal search for new forms, better than the current one, is imperative. That is why evolutionary dynamic systems tend to tune themselves in very particular situations in between regular and chaotic regimes. These particular situations present eternal transients, and the system actually never reaches its final destiny, preserving diversity. This feature allows the system to visit other regions of the space of possibilities, not only the tiny region covered by its final attractor.
The Broad Histogram Method (BHM) allows one to determine the energy degeneracy g(E), i.e. the energy spectrum of a given system, from the knowledge of the microcanonical averages and of two macroscopic quantities Nup and Ndn defined within the method. The fundamental BHM equation relating g(E) to the quoted averages is exact and completely general for any conceivable system. Thus, the only possible source of numerical inaccuracies resides on the measurement of the averages themselves. In this text, we introduce a Monte Carlo recipe to measure microcanonical averages. In order to test its performance, we applied it to the Ising ferromagnet on a 32x32 square lattice. The exact values of g(E) are known up to this lattice size, thus it is a good standard to compare our numerical results with. Measuring the deviations relative to the exactly known values, we verified a decay proportional to 1/sqrt(counts), by increasing the counter (counts) of averaged samples over at least 6 decades. That is why we believe this microcanonical simulator presents no bias besides the normal statistical fluctuations. For counts~10**10, we measured relative deviations near 10**(-5) for both g(E) and the specific heat peak, obtained through BHM relation.
The Broad Histogram is a method allowing the direct calculation of the energy degeneracy $g(E)$. This quantity is independent of thermodynamic concepts such as thermal equilibrium. It only depends on the distribution of allowed (micro) states along the energy axis, but not on the energy changes between the system and its environment. Once one has obtained $g(E)$, no further effort is needed in order to consider different environment conditions, for instance, different temperatures, for the same system. The method is based on the exact relation between $g(E)$ and the microcanonical averages of certain macroscopic quantities $N^{\rm up}$ and $N^{\rm dn}$. For an application to a particular problem, one needs to choose an adequate instrument in order to determine the averages $ $ and $ $, as functions of energy. Replacing the usual fixed-temperature canonical by the fixed-energy microcanonical ensemble, new subtle concepts emerge. The temperature, for instance, is no longer an external parameter controlled by the user, all canonical averages being functions of this parameter. Instead, the microcanonical temperature $T_{m}(E)$ is a function of energy defined from $g(E)$ itself, being thus an {\bf internal} (environment independent) characteristic of the system. Accordingly, all microcanonical averages are functions of $E$. The present text is an overview of the method. Some features of the microcanonical ensemble are also discussed, as well as some clues towards the definition of efficient Monte Carlo microcanonical sampling rules.
The Broad Histogram is a method designed to calculate the energy degeneracy g(E) from microcanonical averages of certain macroscopic quantities Nup and Ndn. These particular quantities are defined within the method, and their averages must be measured at constant energy values, i.e. within the microcanonical ensemble. Monte Carlo simulational methods are used in order to perform these measurements. Here, the mathematical relation allowing one to determine g(E) from these averages is shown to be exact for any statistical model, i.e. any energy spectrum, under completely general conditions. We also comment about some troubles concerning the measurement of the quoted microcanonical averages, when one uses a particular approach, namely the energy random walk dynamics. These troubles appear when movements corresponding to different energy jumps are performed using the same probability, and also when the correlations between successive averaging states are not adequately treated: they have nothing to do with the method itself.