Search

Causal discovery aims to automatically uncover causal relationships from data, a capability with significant potential across many scientific disciplines. However, its real-world applications remain limited. Current methods often rely on unrealistic assumptions and are evaluated only on simple synthetic toy datasets, often with inadequate evaluation metrics. In this paper, we substantiate these claims by performing a systematic review of the recent causal discovery literature. We present applications in biology, neuroscience, and Earth sciences - fields where causal discovery holds promise for addressing key challenges. We highlight available simulated and real-world datasets from these domains and discuss common assumption violations that have spurred the development of new methods. Our goal is to encourage the community to adopt better evaluation practices by utilizing realistic datasets and more adequate metrics., Comment: 39 pages, 8 figures

Scientific research often seeks to understand the causal structure underlying high-level variables in a system. For example, climate scientists study how phenomena, such as El Ni\~no, affect other climate processes at remote locations across the globe. However, scientists typically collect low-level measurements, such as geographically distributed temperature readings. From these, one needs to learn both a mapping to causally-relevant latent variables, such as a high-level representation of the El Ni\~no phenomenon and other processes, as well as the causal model over them. The challenge is that this task, called causal representation learning, is highly underdetermined from observational data alone, requiring other constraints during learning to resolve the indeterminacies. In this work, we consider a temporal model with a sparsity assumption, namely single-parent decoding: each observed low-level variable is only affected by a single latent variable. Such an assumption is reasonable in many scientific applications that require finding groups of low-level variables, such as extracting regions from geographically gridded measurement data in climate research or capturing brain regions from neural activity data. We demonstrate the identifiability of the resulting model and propose a differentiable method, Causal Discovery with Single-parent Decoding (CDSD), that simultaneously learns the underlying latents and a causal graph over them. We assess the validity of our theoretical results using simulated data and showcase the practical validity of our method in an application to real-world data from the climate science field., Comment: 33 pages, 17 figures

Climate models, such as Earth system models (ESMs), are crucial for simulating future climate change based on projected Shared Socioeconomic Pathways (SSP) greenhouse gas emissions scenarios. While ESMs are sophisticated and invaluable, machine learning-based emulators trained on existing simulation data can project additional climate scenarios much faster and are computationally efficient. However, they often lack generalizability and interpretability. This work delves into the potential of causal representation learning, specifically the \emph{Causal Discovery with Single-parent Decoding} (CDSD) method, which could render climate model emulation efficient \textit{and} interpretable. We evaluate CDSD on multiple climate datasets, focusing on emissions, temperature, and precipitation. Our findings shed light on the challenges, limitations, and promise of using CDSD as a stepping stone towards more interpretable and robust climate model emulation.

Climate models have been key for assessing the impact of climate change and simulating future climate scenarios. The machine learning (ML) community has taken an increased interest in supporting climate scientists' efforts on various tasks such as climate model emulation, downscaling, and prediction tasks. Many of those tasks have been addressed on datasets created with single climate models. However, both the climate science and ML communities have suggested that to address those tasks at scale, we need large, consistent, and ML-ready climate model datasets. Here, we introduce ClimateSet, a dataset containing the inputs and outputs of 36 climate models from the Input4MIPs and CMIP6 archives. In addition, we provide a modular dataset pipeline for retrieving and preprocessing additional climate models and scenarios. We showcase the potential of our dataset by using it as a benchmark for ML-based climate model emulation. We gain new insights about the performance and generalization capabilities of the different ML models by analyzing their performance across different climate models. Furthermore, the dataset can be used to train an ML emulator on several climate models instead of just one. Such a "super emulator" can quickly project new climate change scenarios, complementing existing scenarios already provided to policymakers. We believe ClimateSet will create the basis needed for the ML community to tackle climate-related tasks at scale., Comment: To be published in the 37th Conference on Neural Information Processing Systems (NeurIPS 2023): Track on Datasets and Benchmarks. Project website: https://climateset.github.io/

Families of asteroids generated by the collisional fragmentation of a common parent body have been identified using clustering methods of asteroids in their proper orbital element space. An alternative method has been developed in order to identify collisional families from the correlation between the asteroid fragment sizes and their proper semi-major axis distance from the family centre (V-shape). This method has been shown to be effective in the cases of the very diffuse families that formed billions of years ago. We obtained photometric observations of asteroids in order to construct their rotational light curves; we combine them with the literature light curves and sparse-in-time photometry; we input these data in the light curve inversion methods to determine the shape and the spin pole of the asteroids in order to assess whether an object is prograde or retrograde. The ultimate goal is to assess whether we find an excess of retrograde asteroids on the inward side of the V-shape of a 4 Gyr asteroid family identified via the V-shape method. This excess of retrograde rotators is predicted by the theory of asteroid family evolution. We obtained the spin poles for 55 asteroids claimed to belong to a 4 Gyr collisional family of the inner main belt that consists of low-albedo asteroids. After re-evaluating the albedo and spectroscopic information, we found that nine of these asteroids are interlopers in the 4 Gyr family. Of the 46 remaining asteroids, 31 are found to be retrograde and 15 prograde. We also found that these retrograde rotators have a very low probability (1.29%) of being due to random sampling from an underlying uniform distribution of spin poles. Our results constitute corroborating evidence that the asteroids identified as members of a 4 Gyr collisional family have a common origin, thus strengthening their family membership., Comment: 20 pages, 7 figures. A&A (2022)

The ExoClock project has been created with the aim of increasing the efficiency of the Ariel mission. It will achieve this by continuously monitoring and updating the ephemerides of Ariel candidates over an extended period, in order to produce a consistent catalogue of reliable and precise ephemerides. This work presents a homogenous catalogue of updated ephemerides for 450 planets, generated by the integration of $\sim$18000 data points from multiple sources. These sources include observations from ground-based telescopes (ExoClock network and ETD), mid-time values from the literature and light-curves from space telescopes (Kepler/K2 and TESS). With all the above, we manage to collect observations for half of the post-discovery years (median), with data that have a median uncertainty less than one minute. In comparison with literature, the ephemerides generated by the project are more precise and less biased. More than 40\% of the initial literature ephemerides had to be updated to reach the goals of the project, as they were either of low precision or drifting. Moreover, the integrated approach of the project enables both the monitoring of the majority of the Ariel candidates (95\%), and also the identification of missing data. The dedicated ExoClock network effectively supports this task by contributing additional observations when a gap in the data is identified. These results highlight the need for continuous monitoring to increase the observing coverage of the candidate planets. Finally, the extended observing coverage of planets allows us to detect trends (TTVs - Transit Timing Variations) for a sample of 19 planets. All products, data, and codes used in this work are open and accessible to the wider scientific community., Comment: Recommended for publication to ApJS (reviewer's comments implemented). Main body: 13 pages, total: 77 pages, 7 figures, 7 tables. Data available at http://doi.org/10.17605/OSF.IO/P298N

Causal discovery from observational data is a challenging task that can only be solved up to a set of equivalent solutions, called an equivalence class. Such classes, which are often large in size, encode uncertainties about the orientation of some edges in the causal graph. In this work, we propose a new set of assumptions that constrain possible causal relationships based on the nature of variables, thus circumscribing the equivalence class. Namely, we introduce typed directed acyclic graphs, in which variable types are used to determine the validity of causal relationships. We demonstrate, both theoretically and empirically, that the proposed assumptions can result in significant gains in the identification of the causal graph. We also propose causal discovery algorithms that make use of these assumptions and demonstrate their benefits on simulated and pseudo-real data., Comment: 30 pages, 13 figures, accepted for the 1st conference on Causal Learning and Reasoning (CLeaR), 2022

Learning a causal directed acyclic graph from data is a challenging task that involves solving a combinatorial problem for which the solution is not always identifiable. A new line of work reformulates this problem as a continuous constrained optimization one, which is solved via the augmented Lagrangian method. However, most methods based on this idea do not make use of interventional data, which can significantly alleviate identifiability issues. This work constitutes a new step in this direction by proposing a theoretically-grounded method based on neural networks that can leverage interventional data. We illustrate the flexibility of the continuous-constrained framework by taking advantage of expressive neural architectures such as normalizing flows. We show that our approach compares favorably to the state of the art in a variety of settings, including perfect and imperfect interventions for which the targeted nodes may even be unknown., Comment: Appears in: Advances in Neural Information Processing Systems 34 (NeurIPS 2020). 46 pages

Several chemical networks have been developed to study warm (exo)planetary atmospheres. The kinetics of the reactions related to the methanol chemistry included in these schemes have been questioned. The goal of this paper is to update the methanol chemistry for such chemical networks thanks to recent publications in the combustion literature. We aim also at studying the consequences of this update on the atmospheric compositions of (exo)planetary atmospheres and brown dwarfs. We have performed an extensive review of combustion experimental studies and revisited the sub-mechanism describing methanol combustion in the scheme of Venot et al. (2012, A&A 624, A58). The updated scheme involves 108 species linked by a total of 1906 reactions. We have then applied our 1D kinetic model with this new scheme to several case studies (HD 209458b, HD 189733b, GJ 436b, GJ 1214b, ULAS J1335+11, Uranus, Neptune), and compared the results obtained with those obtained with the former scheme. The update of the scheme has a negligible impact on hot Jupiters atmospheres. However, the atmospheric composition of warm Neptunes and brown dwarfs is modified sufficiently to impact observational spectra in the wavelength range JWST will operate. Concerning Uranus and Neptune, the update of the chemical scheme modifies the abundance of CO and thus impacts the deep oxygen abundance required to reproduce the observational data. For future 3D kinetics models, we also derived a reduced scheme containing 44 species and 582 reactions. Chemical schemes should be regularly updated in order to maintain a high level of reliability on the results of kinetic models and be able to improve our knowledge on planetary formation., Comment: 18 pages, 23 figures, accepted for publication in A&A

We propose a novel score-based approach to learning a directed acyclic graph (DAG) from observational data. We adapt a recently proposed continuous constrained optimization formulation to allow for nonlinear relationships between variables using neural networks. This extension allows to model complex interactions while avoiding the combinatorial nature of the problem. In addition to comparing our method to existing continuous optimization methods, we provide missing empirical comparisons to nonlinear greedy search methods. On both synthetic and real-world data sets, this new method outperforms current continuous methods on most tasks, while being competitive with existing greedy search methods on important metrics for causal inference., Comment: Appears in: Proceedings of the Eighth International Conference on Learning Representations (ICLR 2020). 23 pages