Comparison of Performance of Recursive Partitioning Models and Parametric Imputation Models Based On Predictive Mean Matching In Multiple Imputation By Chained Equations, in Order to Impute The Missing Values of The Binary Outcome, in The Presence of An Interaction Effect

Background: Among the new multiple imputation methods, Multiple Imputation by Chained ‎Equations (MICE) is a ‎popular ‎approach for implementing multiple imputations because of its ‎flexibility. Our main focus in this study ‎is to ‎compare the performance of parametric ‎imputation models based on predictive mean matching and ‎recursive partitioning methods ‎in multiple imputation by chained equations in the ‎presence of interaction in the ‎data.Methods: We compared the performance of parametric and tree-based imputation methods via simulation using two data generation models. For each combination of data generation model and imputation method, the following steps were performed: data generation, removal of observations, imputation, logistic regression analysis, and calculation of bias, Coverage Probability (CP), and Confidence Interval (CI) width for each coefficient Furthermore, model-based and empirical SE, and estimated proportion of the variance attributable to the missing data (λ) were calculated.Results: ‎We have shown by simulation that to impute a binary response in ‎observations involving an ‎interaction, manually interring the interaction term into the imputation model in the ‎predictive mean matching ‎model improves the performance of the PMM method compared to the recursive partitioning models in ‎ ‎multiple imputation by chained equations.‎ The parametric method in which we entered the interaction model into the imputation model (MICE-‎‎‎Interaction) led to smaller bias, slightly higher coverage probability for the interaction effect, but it ‎had ‎slightly ‎wider confidence intervals than tree-based imputation (especially classification and ‎regression ‎trees). Conclusions: The application of MICE-Interaction led to better performance than ‎recursive ‎partitioning methods in MICE, although ‎the user is interested in estimating the interaction and does not ‎know ‎enough about the structure of the observations, recursive partitioning methods can be ‎suggested to impute ‎the ‎missing values.

Ecosystem Carbon Fluxes Analyzed Using Eddy Covariance Technique

Eddy covariance (EC) is an important measurement technique used in physical geography and atmospheric sciences to measure the exchange of carbon dioxide between an ecosystem and the atmosphere at a specific location. However, EC produces a net exchange of carbon dioxide yet research questions require an understanding of component fluxes, carbon dioxide uptake by plants through photosynthesis and carbon dioxide emissions due to plant and soil respiration. There are two major methods to partition EC measurements into these component fluxes: night-time and day-time partitioning methods. In the night-time method, nighttime measurements are used to estimate daytime respiration and calculate photosynthesis as a residual and in the daytime method, a light response curve is created to estimate daytime respiration and photosynthesis. This study investigates the benefits and drawbacks of these partitioning methods on two carbon dioxide exchange datasets from ecosystems in Canada. The research sites were a) Mer Bleue, a peatland bog near Ottawa, Ontario and b) Cape Bounty, a high arctic tundra in Nunavut. By using a combination of the REddy-Proc software package, developed by the Max Planck Institute for Biogeochemistry, along with additional Matlab processing, the differences in photosynthesis and respiration due to partitioning methods are presented and discussed.

Influence of partitioning methods on computational cost of cfd simulations applied to hydrocyclones

In simulations, when using multipartitioned computers, the way in which the mesh is partitioned directly affects the average time per iteration, and therefore, the computational cost. This study proposes an analysis of the average time per iteration of 23 different partition methods available for tridimensional mesh in software FLUENT 19.2. For the calculation of the average iteration time, 100 iterations were used. Generally, the best partitioning methods were those in which the mesh division was made perpendicularly to the axis of the equipment. It was stated the choice of an adequate partitioning method can save high costs of computational power. For the hydrocyclone studied, with a computer with 8 cores, approximately 24.56 hours of simulation were saved, representing almost 20% of the total time.