DOC update index, readme and quantile regression example (#94)

* DOC update index, readme and quantile regression example * CLN remove duplicate import and fix typos * DOC fix more typos
lorentzenchr · Jul 16, 2023 · b868ff4 · b868ff4
1 parent 48f9abc
commit b868ff4
Show file tree

Hide file tree

Showing 4 changed files with 104 additions and 53 deletions.
diff --git a/README.md b/README.md
@@ -12,15 +12,15 @@
 Highlights:
 
 - All common point predictions covered: mean, median, quantiles, expectiles.
-- Assess model calibration with identification functions (generalized residuals).
+- Assess model calibration with [identification functions](https://lorentzenchr.github.io/model-diagnostics/reference/model_diagnostics/calibration/identification/#model_diagnostics.calibration.identification.identification_function) (generalized residuals) and [compute_bias](https://lorentzenchr.github.io/model-diagnostics/reference/model_diagnostics/calibration/identification/#model_diagnostics.calibration.identification.compute_bias).
 - Assess calibration and bias graphically
-    - reliability diagrams for auto-calibration
-    - bias plots for conditional calibration
+    - [reliability diagrams](https://lorentzenchr.github.io/model-diagnostics/reference/model_diagnostics/calibration/plots/#model_diagnostics.calibration.plots.plot_reliability_diagram) for auto-calibration
+    - [bias plots](https://lorentzenchr.github.io/model-diagnostics/reference/model_diagnostics/calibration/plots/#model_diagnostics.calibration.plots.plot_bias) for conditional calibration
 - Assess the predictive performance of models
-    - strictly consistent, homogeneous scoring functions
-    - score decomposition into miscalibration, discrimination and uncertainty
+    - strictly consistent, homogeneous [scoring functions](https://lorentzenchr.github.io/model-diagnostics/reference/model_diagnostics/scoring/scoring/)
+    - [score decomposition](https://lorentzenchr.github.io/model-diagnostics/reference/model_diagnostics/scoring/scoring/#model_diagnostics.scoring.scoring.decompose) into miscalibration, discrimination and uncertainty
 
-:rocket: To our knowledge, this is the first python package to offer reliability diagrams for quantiles and expectiles made available by an internal implementation of isotonic quantile/expectile regression. :rocket:
+:rocket: To our knowledge, this is the first python package to offer reliability diagrams for quantiles and expectiles and a score decomposition, both made available by an internal implementation of isotonic quantile/expectile regression. :rocket:
 
 Read more in the [documentation](https://lorentzenchr.github.io/model-diagnostics/).
 

diff --git a/docs/examples/quantile_regression.ipynb b/docs/examples/quantile_regression.ipynb
diff --git a/docs/examples/regression_on_workers_compensation.ipynb b/docs/examples/regression_on_workers_compensation.ipynb
@@ -492,7 +492,7 @@
    "source": [
     "### 2.2 OLS\n",
     "\n",
-    "Here, we train a Ordinary Least Squares (OLS) model, but on the log transformed target.\n",
+    "Here, we train an Ordinary Least Squares (OLS) model, but on the log transformed target.\n",
     "`TransformedTargetRegressor` takes care of this transformation and also of the back transformation such that predictions are on the original target UltimateIncurredClaimCost."
    ]
   },
@@ -836,7 +836,7 @@
    "source": [
     "For a better visual comparison, we use the `diagram_type=\"bias\"` option to rotate the plots by 45 degrees (and mirror on the horizontal line) and plot all models in one go.\n",
     "This means, closer to y=0 is better.\n",
-    "On top, we set the x-axis to a logarithmic scale and we zoom in a little."
+    "On top, we set the x-axis to a logarithmic scale and zoom in a little."
    ]
   },
   {
@@ -1186,7 +1186,7 @@
     "The deciding difference of the scores then stems from the **miscalibration** term.\n",
     "Except for the trivial model, the HGBT model has the best (smallest) one.\n",
     "The **discrimination** term (larger is better) is quite similar among the non-trivial models.\n",
-    "Interestingly, the Gamma GLM has a better miscalibration term but a worse discrimination term compared to the Poisson GLM."
+    "Interestingly, the Gamma GLM has a better miscalibration term, but a worse discrimination term compared to the Poisson GLM."
    ]
   },
   {

diff --git a/docs/index.md b/docs/index.md
@@ -12,15 +12,15 @@
 Highlights:
 
 - All common point predictions covered: mean, median, quantiles, expectiles.
-- Assess model calibration with [identification functions][model_diagnostics.calibration.identification.identification_function] (generalized residuals).
+- Assess model calibration with [identification functions][model_diagnostics.calibration.identification.identification_function] (generalized residuals) and [compute_bias][model_diagnostics.calibration.identification.compute_bias].
 - Assess calibration and bias graphically
     - [reliability diagrams][model_diagnostics.calibration.plots.plot_reliability_diagram] for auto-calibration
     - [bias plots][model_diagnostics.calibration.plots.plot_bias] for conditional calibration
 - Assess the predictive performance of models
-    - strictly consistent, homogeneous scoring functions
+    - strictly consistent, homogeneous [scoring functions][model_diagnostics.scoring.scoring]
     - [score decomposition][model_diagnostics.scoring.decompose] into miscalibration, discrimination and uncertainty
 
-:rocket: To our knowledge, this is the first python package to offer reliability diagrams for quantiles and expectiles made available by an internal implementation of isotonic quantile/expectile regression. :rocket:
+:rocket: To our knowledge, this is the first python package to offer reliability diagrams for quantiles and expectiles and a score decomposition, both made available by an internal implementation of isotonic quantile/expectile regression. :rocket:
 
 This package relies on the giant shoulders of, among others, [polars](https://pola.rs/), [matplotlib](https://matplotlib.org), [scipy](https://scipy.org) and [scikit-learn](https://scikit-learn.org).