• Energy Research
• Research software close

MASSIMO is a distance-independent individual-tree simulator that represents demographic processes (regeneration, growth and mortality) with empirical models that have been parameterized with data from the Swiss NFI. Tree regeneration, growth and mortality are simulated on the regular grid of sample plots of the Swiss NFI, which allows for statistically representative simulations of forest development. 

Analysis and calibration of optimal power balance rotor-effective wind speed estimation schemes for large-scale wind turbines – Code and Data This repository contains the MATLAB scripts and Simulink models associated with the paper: Pamososuryo, A. K., Spagnolo, F., Mulders, S. P. Analysis and calibration of optimal power balance rotor-effective wind speed estimation schemes for large-scale wind turbines Wind Energy Science, 2024 – Preprint available here DOI: 10.5194/wes-2024-158 Repository Purpose This archive reproduces all computational results, figures, and numerical validations presented in the manuscript. The code provides a fully operational framework for simulating, analyzing, and comparing rotor-effective wind speed (REWS) estimators based on the power balance principle. Contents PowerBalanceLOFI/ Contains low-fidelity studies, estimator calibration, and robustness analysis: Script1_WindTurbineDataCurveFitting.m: Curve fitting for turbine property scaling (inertia, rated power). Script2_FilteredDerivativeNoiseStudy.m: Evaluates numerical derivative sensitivity under noisy conditions. Script3_LuenbergerAeroPowerEstimator.m: Implements a state-estimation-based aerodynamic power estimator. Script4_EstimatorSolverComparison.m: Compares solver strategies for the REWS solver component. Script5_ContinuousSolverStability.m: Investigates stability of continuous solvers under sampling effects. .slx models: Matching Simulink files for each scenario above. PowerBalanceOpenFAST/ Includes the high-fidelity validation setup using OpenFAST: Script1_main_OpenFAST.m: Runs OpenFAST simulations to generate rotor dynamics data. Script2_OpenLoopEstimation.m: Executes the power balance wind speed estimator (PB-WSE) using measured signals. Script3_Plotting.m: Produces time series and histogram figures for REWS estimation analysis. OpenFAST.slx, OpenLoopEstimation.slx: Simulink models implementing the PB-WSE structure. dependencies/ The dependencies/ folder contains third-party functions used for plotting and figure export: export_fig/: External tool for exporting figures with high quality and transparency. Source: https://github.com/altmany/export_fig linspecer/: Color palettes for line plots with distinguishable colors. Source: MathWorks File Exchange matplotlib/: MATLAB-based colormaps mimicking Python’s matplotlib perceptually uniform colormaps. Source: MathWorks File Exchange setfigpaper/: Utility to standardize figure layout and export style. Source: https://github.com/jmrplens/SetFigPaper Estimation Architecture The proposed estimator is split into two calibrated modules: Aerodynamic Power Estimator Based on either: * Numerical derivative of rotor speed * Luenberger observer for aerodynamic torque estimation Wind Speed Estimate Solver Implemented as: * Continuous-time solver * Iterative single-step solver The final configuration—state-estimation-based aerodynamic power estimator + iterative solver—is shown to be optimal. Requirements MATLAB R2024b or newer Simulink Curve Fitting Toolbox OpenFAST 3.5.3 MATLAB/Simulink interface Citation Please cite the following work when using this repository: A. K. Pamososuryo, F. Spagnolo, S. P. Mulders Analysis and calibration of optimal power balance rotor-effective wind speed estimation schemes for large-scale wind turbines Wind Energy Science, 2024 DOI: 10.5194/wes-2024-158 Authors & Affiliations Atindriyo K. Pamososuryo, Delft Center for Systems and Control, TU Delft Fabio Spagnolo, Vestas Wind Systems A/S Sebastiaan P. Mulders, Delft Center for Systems and Control, TU Delft Contact Corresponding author: A.K.Pamososuryo@tudelft.nl

Runoff and water temperature data We estimated mean annual precipitation, averaged across each drainage area, using Google Climate Engine, March 2011 - February 2021. Where multiple temperature loggers were present in a study stream, we selected a single location based on the completeness of data in the study season and proximity to the PIT antenna. Hourly temperature measurements were converted into mean daily values. Analysis For data analysis and modeling, we excluded streams that had less than 3 years of biological data, leaving 47 stream-years. We conducted all analyses in R (version 4.0.4, R Core Team, 2018). We tested outmigration timing data for normal distribution among streams, years, and stream-years, using the shapiro.test function of the broom package. The Shapiro-Wilk test showed that all distributions were unlikely to be normally distributed (i.e. among years, p = 5.5 × 10-9 – 7.6 × 10-39 and W = 0.88 – 0.98). However, the Shapiro-Wilk test can provide small p-values for large samples and consequently provide a false negative, regarding normal distribution (among years, sample size range was 485 – 3453). Therefore, we could not rule out the possibility of assumptions being met to perform ANOVA (Analysis of Variance) tests. We did so, using the aov function of the AICcmodavg package: i) one-way, by stream, ii) a one-way, by year, iii) a two-way, by stream and year, and iv) a two-way with stream-year interaction. To isolate the effects of stream and year on variance, we performed the ANOVA tests on the maximum subset of data for which each stream had the same years of outmigration (four streams, each with the same six years of data, totaling 24 stream-years). The aictab function of the AICcmodavg package demonstrated that the two-way model with stream-year interaction was the highest performing (lowest AICc value), followed by: the two-way model, one-way by year model, and one-way by stream model. In both ANOVA tests, the year, stream, and year-stream interaction terms each had "Pr(>F)" values < 2 × 10-16. The "2-way ANOVA with interaction" (year F-value 646.58, stream F-value 349.85, year-stream interaction F-value 29.31, residuals 4.11 × 10-16) had higher F values and lower residuals than the 2-way ANOVA (year F-value 629.3, stream F-value 340.5, residuals 4.22 × 10-16). We used the TukeyHSD function of the AICcmodavg package to conduct pairwise tests for significant differences in outmigration timing distributions. Among streams, five of six pairwise differences were highly significant (p < 0.0001). Among years, all 15 pairwise comparisons were highly significant (p < 0.001). Among stream-years, 216 of 277 pair-wise comparisons were significant (p < 0.05). We checked for homoscedasticity in the interaction model, using the leveneTest function of the car library, and we found evidence that the variance across groups is significantly different. Consequently, we cannot assume homogeneity of variances in the different groups, which is typically a required assumption for conducting ANOVA tests. Since the normal distribution assumption of the one-way ANOVA was not met, we applied the Kruskal-Wallis test, as a non-parametric alternative to test for variance among streams and years, using the package rstatix. As with the ANOVA tests, we performed Kruskal-Wallis tests on the maximum subset of data for which each stream had the same years of outmigration (24 stream-years), using the functions kruskal_test, kruskal_effsize, dunn_test, and wilcox_test. Among streams, we found significant variance (p = 2.16 × 10-143), with a "small" effect size (eta-squared measure = 0.04) (Tomczak and Tomczak 2014), and 5 of 6 pairwise differences were highly significant (Dunn's test & Wilcoxon's test: p < 0.0001). Among years, we found significant variance (p = 0), with a "large" effect size (eta-squared measure = 0.17) (Tomczak and Tomczak 2014), and 13 of 15 pairwise differences were highly significant (Dunn's test & Wilcoxon's test: p < 0.0001). Modeling the effects of streamflow and water temperature on outmigration timing Modeling was limited to the 42 stream-years for which water temperature and outmigration timing data were collected. For the outmigration start date model, the runoff date range was March-April and the degree-days date range was March-April. For the outmigration end date and duration models, the runoff date range was March-June and the degree-days date range was March-April. Coefficient units are "days per daily runoff (mm)" and "days per 100 degree-days". In identifying top model(s), we did not consider degree-days to influence outmigration duration because: i) the AIC value of the runoff-only model was 1.99 less than the additive model, ii) the degree-days in the additive model had a p-value > 0.05, and iii) Mar-Jun runoff had similar coefficient effect sizes in the additive model and run-off only model (Appendix S1: Table S3). We calculated conditional coefficients (including stream, as a random effect) and marginal coefficients (excluding stream, as a random effect) of determination (R2) (Nakagawa and Schielzeth 2013), using the r.squaredGLMM function of the MuMIn package (Barton` 2020). We also reported the model coefficients and 95% confidence intervals, as measures of effect size, and generated partial dependence plots for using the plot_model function of the sjPlot package (Lüdecke 2021). Literature cited Barton`, K. (2020). MuMIn: Multi-Model Inference. R package version 1.43.17. Lüdecke, D. (2021). sjPlot: Data Visualization for Statistics in Social Science. R package version 2.8.9. Nakagawa, S., and H. Schielzeth. 2013. A general and simple method for obtaining R 2 from generalized linear mixed-effects models. Methods in Ecology and Evolution 4:133–142. Tomczak, M., and E. Tomczak. 2014. The need to report effect size estimates revisited. An overview of some recommended measures of effect size 1:7. Prolonged migration windows buffer migratory animal populations against uncertainty in resource availability. Understanding how intensifying droughts from climate change influence the migration window is critical for biodiversity conservation in a warming world. We explored how drought affects the seaward migration of endangered coho salmon (Oncorhynchus kisutch) near the southern extent of their range in California, USA. We tracked stream departures of juvenile coho, measuring streamflow and temperature in 7 streams over 13 years, spanning an historic drought with extreme dry and warm conditions. Linear mixed effects models indicate that, over the range of observations, a decrease in seasonal streamflow (from 4.5 to 0.5 mm/day seasonal runoff) contracted the migration window by 31% (from 11 to 7 weeks). An increase from 10.2 to 12.8 ℃ in mean seasonal water temperature hastened the migration window by three weeks. Pacific salmon have evolved to synchronize ocean arrival with productive ocean upwelling. However, earlier and shorter migration windows during drought could lead to mismatches, decreasing fitness and population stability. Our study demonstrates that drought-induced low flows and warming threaten coho salmon in California and suggests that environmental flow protections will be needed to support the seaward migration of Pacific salmon in a changing climate. Please see DataS1/data/README_Metadata.pdf.Funding provided by: California Department of Fish and WildlifeCrossref Funder Registry ID: http://dx.doi.org/10.13039/100006238Award Number: Funding provided by: California Sea Grant, University of California, San DiegoCrossref Funder Registry ID: http://dx.doi.org/10.13039/100005522Award Number: Graduate Research Fellowship R/AQ-153FFunding provided by: National Geographic SocietyCrossref Funder Registry ID: http://dx.doi.org/10.13039/100006363Award Number: EC-53369R-18Funding provided by: National Oceanic and Atmospheric AdministrationCrossref Funder Registry ID: http://dx.doi.org/10.13039/100000192Award Number: Funding provided by: National Science FoundationCrossref Funder Registry ID: http://dx.doi.org/10.13039/100000001Award Number: Graduate Research Fellowship DGE 1752814Funding provided by: Sonoma Fish and Wildlife Commission*Crossref Funder Registry ID: Award Number: Funding provided by: U.S. Army Corps of EngineersCrossref Funder Registry ID: http://dx.doi.org/10.13039/100006752Award Number:

The tool presented here is a BETA version, bringing together most of the tool functionalities discussed during the partner workshops. The aim of this version is to provide an overview of developments since the start of the Cimopolée project. It also enables users to report anomalies (comments on improvements, bugs, desired modifications, etc.) via an anomaly report..

GERALDINE is a free-to-use resource that enables the detection and characterisation of mass movements onto glaciers. Tool available at: GERALDINE (v1.1) Citation: Smith, W. D., Dunning, S. A., Brough, S., Ross, N., and Telling, J.: GERALDINE (Google Earth Engine supRaglAciaL Debris INput dEtector): a new tool for identifying and monitoring supraglacial landslide inputs, Earth Surf. Dynam., 8, 1053–1065, https://doi.org/10.5194/esurf-8-1053-2020, 2020. Version 1.1 removes the NDWI mask from the GERALDINE processing flow following reviewer comments.

This archive contains all relevant information to reproduce the experimental figures presented in the paper "A 5/4(1+eps)-Approximation Algorithm for Scheduling with Rejection Costs Proportional to Processing Times", as well as the scripts to re-run those experiments and new ones and process the results. All results presented in the paper are archived here as well.

The macroscale at which we measure, model, and predict climate change does not align with the microscale at which small ectotherms experience climate. To understand climate's influence on biodiversity and potential ecological effects of climate change, more work is needed to understand how ectotherm physiology relates to microclimatic temperatures. Tree canopies are an example of a habitat that produces extreme microclimates, and arthropods in tropical forest canopies are threatened by extreme heat and warming. The situation in temperate canopies, however, is less clear. Conventional wisdom suggests that winter cold limits arboreal arthropod diversity in temperate forests, but because the canopy is less buffered from extreme temperatures, summer heat could also play a role. Heat- and cold-limited communities will respond differently to climate change, so this distinction is critical. Using the frameworks of the thermal adaptation hypothesis and thermal niche asymmetry, we asked whether arboreal ants were physiologically adapted to their extreme environment and whether summer heat or winter cold was more stressful. We tracked internal microclimates of ant nests in the canopy and on the ground over the seasonal cycle in temperate forests in North Carolina, USA. Then, we measured the heat (CTmax) and cold tolerance (CTmin) of worker ants in summer and spring and compared them to the ants' experienced microclimates. Nests in the temperate canopy experienced hotter and colder extremes and more closely tracked air temperatures than ant nests on the ground. Arboreal ants partially adhered to the thermal adaptation hypothesis. They were more heat-tolerant than ground-nesting species, but despite experiencing lower temperatures, they were less cold-tolerant. Ants acclimated their cold tolerance in line with seasonal changes, but heat tolerance was more phylogenetically constrained. Summer heat did not approach ants' heat tolerance in either stratum, but winter and spring lows in the canopy exceeded the cold tolerance of ants nesting there. By comparing microclimatic temperatures and thermal physiology, we show that winter cold—and not summer heat—likely limits arthropod diversity in the temperate canopy. As the climate warms, the temperate canopy may become accessible to more arthropod species. Funding provided by: North Carolina State UniversityROR ID: https://ror.org/04tj63d06Award Number:

This repository contains the replication package for the paper "The EarlyBIRD Catches the Bug: On Exploiting Early Layers of Encoder Models for More Efficient Code Classification" by Anastasiia Grishina, Max Hort and Leon Moonen, accepted for publication in the ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE 2023). The paper is deposited on arXiv and will be available under open access at the publisher's site (IEEE). The replication package is archived on Zenodo with DOI: 10.5281/zenodo.7608802. The source code is distributed under the MIT license, the data is distributed under the CC BY 4.0 license. Citation If you build on this data or code, please cite this work by referring to the paper: @inproceedings{grishina2023:earlybird, title = {The EarlyBIRD Catches the Bug: On Exploiting Early Layers of Encoder Models for More Efficient Code Classification}, author = {Anastasiia Grishina and Max Hort and Leon Moonen}, booktitle = {ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE)}, year = {2023}, publisher = {ACM}, doi = {https://doi.org/10.1145/3611643.3616304}, note = {Pre-print on arXiv at https://arxiv.org/abs/2305.04940} } Organization The replication package is organized as follows: src - the source code requirements - txt files with Python packages and versions for replication data - all raw datasets used for training raw devign - Devign reveal - ReVeal break_it_fix_it - BIFI dataset exception - Exception Type dataset mlruns - results of experiments, the folder is created once the run.py is executed (see part II), empty folder at the time of distribution output - results of experiments tables mlflow_<dataset_name>.csv - we used MLflow to log metrics and parameters in our experiments and generated .csv files with the mlflow experiments csv -x <experiment_number> -o mlflow_<dataset_name>.csv command figures - figures reported in paper runs - folder to store model checkpoints, if the corresponding argument is provided when running the code model-checkpoints - models with the best F1-weighted score on each of the four datasets - one model for one dataset. Note that the best model is not always the model with the best average improvement over the baseline reported in the paper, because of possible best-performing outliers. This folder is distributed as a separate file called EarlyBIRD_model-checkpoints.zip (~4.5GB). notebooks - one Jupyter notebook with code to generate figures and tables with aggregated results as reported in the paper Usage Python version: 3.7.9 (later versions should also work well); CUDA version: 11.6; Git LFS. Commands below work well on Mac or Linux and should be adapted if you have a Windows machine. I. Set up data, environment and code 1. Path to project directory Update path/to/project to point at EarlyBIRD export EarlyBIRD=~/path/to/EarlyBIRD 2. Download codebert checkpoint Please, install Git LFS: https://docs.github.com/en/repositories/working-with-files/managing-large-files/installing-git-large-file-storage Run the following from within $EarlyBIRD/: cd $EarlyBIRD mkdir -p checkpoints/reused/model cd checkpoints/reused/model git lfs install git clone https://huggingface.co/microsoft/codebert-base cd codebert-base/ git lfs pull cd ../../.. 3. Set up a virtual environment cd $EarlyBIRD python -m venv venv source venv/bin/activate 3.1 No CUDA python -m pip install -r requirements/requirements_no_cuda.txt 3.2 With CUDA (to run on GPU) python -m pip install -r requirements/requirements_with_cuda.txt python -m pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu116 4 Preprocess data After preprocessing, all datasets are stored in jsonlines (if in python) format. Naming convention: split is one of 'train', 'valid', 'test' in data/preprocessed-final/<dataset_name>/<split>.jsonl, with {'src': "def function_1() ...", 'label': "Label1"} {'src': "def function_2() ...", 'label': "Label2"} ... 4.1 Devign Raw data is downloaded from https://drive.google.com/file/d/1x6hoF7G-tSYxg8AFybggypLZgMGDNHfF/view. Test, train, valid txt files are downloaded from the https://github.com/microsoft/CodeXGLUE/tree/main/Code-Code/Defect-detection/ dataset. All files are saved in data/raw/devign. To preprocess raw data: cd $EarlyBIRD python -m src.preprocess \ --dataset_name devign \ --shrink_code \ --config_path src/config.yaml 4.2 ReVeal Raw data is downloaded from https://github.com/VulDetProject/ReVeal under "Our Collected vulnerabilities from Chrome and Debian issue trackers (Often referred as Chrome+Debian or Verum dataset in this project)" and saved in data/raw/reveal. To preprocess raw data: cd $EarlyBIRD python -m src.preprocess \ --dataset_name reveal \ --shrink_code \ --config_path src/config.yaml 4.3 Break-it-fix-it Raw data is downloaded as data_minimal.zip from https://github.com/michiyasunaga/BIFI under p. 1, unzipped, and the folder orig_bad_code is saved in data/raw/break_it_fix_it. To preprocess raw data: cd $EarlyBIRD python -m src.preprocess \ --dataset_name break_it_fix_it \ --shrink_code \ --ratio_train 0.9 \ --config_path src/config.yaml Note: The original paper contains only train and test split. Use --ratio_train to specify what part of the original train (orig-train) split will be used in train and the rest of orig-train will be used for validation during training. 4.4 Exception Type Raw data is downloaded from https://github.com/google-research/google-research/tree/master/cubert under "2. Exception classification" (it points to this storage) and saved in data/raw/exception_type. To preprocess raw data: cd $EarlyBIRD python -m src.preprocess \ --dataset_name exception \ --shrink_code \ --config_path src/config.yaml II. Run code Activate virtual environment (if not done so yet): cd $EarlyBIRD source venv/bin/activate Example run Run experiments with Devign using pruned models (cutoff_layers_one_layer_cls) to 3 layers (--hidden_layer_to_use 3), for example: cd $EarlyBIRD python -m src.run --help # for help with command line args python -m src.run \ --config_path src/config.yaml \ --model_name codebert \ --model_path "checkpoints/reused/model/codebert-base" \ --tokenizer_path "checkpoints/reused/model/codebert-base" \ --dataset_name devign \ --benchmark_name acc \ --train \ --test \ -warmup 0 \ --device cuda \ --epochs 10 \ -clf one_linear_layer \ --combination_type cutoff_layers_one_layer_cls \ --hidden_layer_to_use 3 \ --experiment_no 12 \ --seed 42 To run experiments on a small subset of data, use --debug argument. For example: python -m src.run \ --debug \ --config_path src/config.yaml \ --model_name codebert \ --model_path "checkpoints/reused/model/codebert-base" \ --tokenizer_path "checkpoints/reused/model/codebert-base" \ --dataset_name devign \ --benchmark_name acc \ --train \ --test \ -warmup 0 \ --device cuda \ --epochs 2 \ -clf one_linear_layer \ --combination_type cutoff_layers_one_layer_cls \ --hidden_layer_to_use 3 \ --experiment_no 12 \ --seed 42 Explore output Your EarlyBIRD/ should contain mlruns/. If you started the run.py from another location, you will find mlruns/one level below that location. cd $EarlyBIRD mlflow ui Alternatively, find tables in EarlyBIRD/output/tables/ with best epoch logs and logs of all epochs. ChangeLog v1.0 - corresponds to the version submitted for review to ESEC/FSE 2023 and contains code for using CodeBERT as a base model for fine-tuning, extensive logging in MLFlow and a custom table, as well as replication instructions. v1.1 - corresponds to the camera-ready submission for ESEC/FSE 2023 and contains the code with configurations adapted to use more models for fine-tuning, logging in MLFlow (redundant logging in a custom table is removed), Jupyter notebooks to replicate artifacts in the paper, as well as replication instructions and model checkpoints. Acknowledgement The work included in this repository was supported by the Research Council of Norway through the secureIT project (IKTPLUSS #288787). Max Hort is supported through the ERCIM 'Alain Bensoussan' Fellowship Programme. The empirical evaluation was performed on the Experimental Infrastructure for Exploration of Exascale Computing (eX3), financially supported by the Research Council of Norway under contract #270053, as well as on resources provided by Sigma2, the National Infrastructure for High Performance Computing and Data Storage in Norway.

ERIGrid JRA2: Test case TC3 mosaik implementation