The question | A fixture with known structure | The short answer | The profile | The four criteria | Choosing, and the flat region | What the ladder can support | Choosing on one half, estimating on the other | Next

Why featR | One convention | One return type | A supervised filter | Removing redundancy | Information gain | Methods that need a modeling engine | Choosing a method | What featR will not do quietly

Four questions a score does not answer | A fixture and a model that is wrong in a known way | 1. Structure left in the residuals | 2. Standard errors on aggregated values | 3. Is the cell mean the best estimate of the cell? | 4. Where the score applies | 5. Two ways to leak | Next

What this package is for | Installing | Getting your data in | A spatial file | A table with coordinate columns | What the numbers are in | The pipeline, end to end | Regions and aggregates | Cells or points? | A model, scored honestly | A surface, and where to trust it | Where to go next | Words that recur

The problem with a random fold | A learner you can read | The size of the effect | The five fold schemes | How wide should a block be? | Sizing it by measurement | Reading the result | Every row and every fold is accounted for | Calibration, for a model with intervals | Next

Overview | 1. Packages and boundary | 2. Synthetic observations | 3. Four tessellations | 3a. Voronoi from ~40 k-means seeds | 3b and 3c. Hex and square grids (~50 cells) | 3d. Delaunay triangles | 4. Cell-level aggregation with summarize_by_cell() | Choropleths | 5. Spatial cross-validation | 5a. How far does correlation reach? | 5b. Two fold schemes on the same data | 5c. The number the fold scheme changes | 6. Fitting a model | Residual diagnostics | GWR, if GWmodel is installed | Comparing backends on identical folds | 7. From a fit to a map, and where the map applies | Summary | Where to go next

The regions as a file someone else can use | Grouping a layer you already have | Is this location inside? | IDs that survive a reprojection | Writing it out | The numbers the run already recorded | A worked report | What each line is for | What to save | Next | References