Package 'dengueriskcdmx'

Aedes aegypti ovitrap data, 2017-2024 (full database)

Description

Ovitrap presences and pseudo-absences used to fit the Ae. aegypti occurrence models (primary scenario: no environmental filter on pseudo-absences, 2 km buffer, no circular predictors). The split column records which rows belong to the training and to the test partition (see ae_aegypti_2017_2024_train and ae_aegypti_2017_2024_test).

Usage

ae_aegypti_2017_2024

Format

An sf object with one row per point and the columns:

class

Factor: presence or pseudoabs.

built

Built-up area probability (Dynamic World V1).

tree

Tree-cover probability (Dynamic World V1).

ndvi

NDVI (Sentinel-2, complemented with Landsat 8 for 2017).

prcp

Annual precipitation (MexHiResClimDB).

ia

CONAPO Accessibility Index, interpolated with SPDE-INLA.

ice

CONAPO Environmental Quality Index, interpolated with SPDE-INLA.

im

CONAPO Marginalization Index, interpolated with SPDE-INLA.

pop

Population (LandScan Global).

hfp

Human footprint.

split

Partition: train or test.

geometry

Point geometry (EPSG:4326).

Details

All predictors were harmonised to a common grid of about 500 m. The partition is 80/20, stratified by class, with set.seed(12345) (rsample::initial_split(prop = 0.8, strata = class)).

Source

Ovitrap surveillance of the Mexican Ministry of Health (CENAPRECE); predictors as listed above.

---

Aedes aegypti test partition, 2017-2024

Description

The 20% of ae_aegypti_2017_2024 held out to test the models.

Usage

ae_aegypti_2017_2024_test

Format

An sf object with the same columns as ae_aegypti_2017_2024_train.

Details

All predictors were harmonised to a common grid of about 500 m. The partition is 80/20, stratified by class, with set.seed(12345) (rsample::initial_split(prop = 0.8, strata = class)).

---

Aedes aegypti training partition, 2017-2024

Description

The 80% of ae_aegypti_2017_2024 used to train the models.

Usage

ae_aegypti_2017_2024_train

Format

An sf object with the same columns as ae_aegypti_2017_2024 (except split).

Details

All predictors were harmonised to a common grid of about 500 m. The partition is 80/20, stratified by class, with set.seed(12345) (rsample::initial_split(prop = 0.8, strata = class)).

---

Aedes aegypti external validation data, 2025

Description

Ovitrap presences and pseudo-absences from 2025, not used to fit or tune the models. Used for the external validation. Temperature and elevation, which are not predictors of the primary scenario, are not included.

Usage

ae_aegypti_2025

Format

An sf object with the same columns as ae_aegypti_2017_2024_train.

---

Read a raster shipped with the package

Description

Read a raster shipped with the package

Usage

ae_raster(layer = c("aedes", "dengue", "risk"))

Arguments

layer

One of "aedes" (probability of Aedes aegypti occurrence), "dengue" (probability of confirmed imported dengue cases) or "risk" (five-level epidemiological risk).

Value

A terra::SpatRaster of about 500 m. "aedes" and "dengue" hold the probability (0-1), the mean of the Random Forest, LightGBM and XGBoost models. "risk" has two layers: cluster (k-means cluster) and risk (the five-level epidemiological risk).

Examples

## Not run: 
terra::plot(ae_raster("aedes"))

## End(Not run)

---

Path to a raster shipped with the package

Description

Path to a raster shipped with the package

Usage

ae_raster_path(layer = c("aedes", "dengue", "risk"))

Arguments

layer

One of "aedes" (probability of Aedes aegypti occurrence), "dengue" (probability of confirmed imported dengue cases) or "risk" (five-level epidemiological risk).

Value

Character path to the GeoTIFF in inst/extdata.

Examples

ae_raster_path("risk")

---

Study area

Description

Polygon of the study area (Mexico City Metropolitan Area) used to crop the predictors and the risk maps.

Usage

aoi

Format

An sf object with one feature (POLYGON, WGS 84) and the columns id and geometry.

---