How it works | The aggregation mechanics, without the network | A synthetic two-region country | The sea-level equivalent

Building a synthetic ERA5-like extract | Temperature: temperature_component() | Precipitation: precipitation_component() | Drought: drought_component() | Wind: wind_component() | Sea level: sealevel_component() | Caching heavy computations (steps 3-4)

What is the Actuarial Climate Index? | The four-step workflow | Where things get cached | A minimal worked example

Why a second engine? | Same inputs, same outputs | Via calculate_aci() | The cores argument

Building one small synthetic dataset | Step 1 — compute every component once | Pre-seeding the administrative cache (no network) | The three spatial modes, one temporal granularity | The four granularities, one spatial mode | The full matrix | Mapping one of the grid-cell combinations | Wrap-up

Preparing example results | Time series: plot_aci_timeseries() | Component breakdown: plot_aci_components() | Distributions: plot_aci_distribution() | Maps: plot_aci_map() | Dashboard: plot_aci_dashboard() | Animated maps: animate_aci_map()

Building the synthetic dataset | Running the pipeline | Reading the output | Changing granularity without recomputing | Grid-cell mode: output for mapping

Estimation of the unknown component weight in an admixture model | The one-sample case | Case of symmetric unknown density | Other cases | The two-sample case | Under the null hypothesis $H_0: F_1 = F_2$ | Under the alternative hypothesis $H_1: F_1 \neq F_2$ | Estimation of the unknown cumulative distribution function | References

Algorithm | Applications | On $\mathbb{R}^+$

The one-sample case only available to symmetric unknown density | The two-sample case | Case of symmetric unknown densities | Case of fully unknown densities | The K-sample case | References