Doc: Fix some typos in model selection example (#1140)

and remove unused imports.

doc/example/model_selection.ipynb

@@ -57,7 +57,7 @@
     "\n",
     "| model_subspace_id | petab_yaml                             | $\\theta_1$ | $\\theta_2$ | $\\theta_3$ |\n",
     "|:---------|:----------------------------------|:----|:----|:----|\n",
-    "| M1_0\t| example_modelSelection.yaml\t| 0\t  | 0   |\t0 | \n",
+    "| M1_0\t| example_modelSelection.yaml\t| 0\t  | 0   |\t0 |\n",
     "| M1_1\t| example_modelSelection.yaml\t| 0\t  | 0\t| estimate |\n",
     "| M1_2\t| example_modelSelection.yaml\t| 0\t  | estimate |\t0 |\n",
     "| M1_3\t| example_modelSelection.yaml\t| estimate |\t0\t| 0 |\n",
@@ -249,7 +249,6 @@
     "import pypesto.logging\n",
     "\n",
     "pypesto.logging.log(level=logging.WARNING, name=\"pypesto.petab\", console=True)\n",
-    "import petab\n",
     "\n",
     "pypesto_select_problem_1 = pypesto.select.Problem(\n",
     "    petab_select_problem=petab_select_problem\n",
@@ -260,7 +259,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Models can be selected with a model selection algorithm (here: [forward](https://en.wikipedia.org/wiki/Stepwise_regression)) and a comparison criterion (here: [AIC](https://en.wikipedia.org/wiki/Akaike_information_criterion)). The forward method start with the smallest model. Within each following iteration it tests all models with one additional estimated parameter.\n",
+    "Models can be selected with a model selection algorithm (here: [forward](https://en.wikipedia.org/wiki/Stepwise_regression)) and a comparison criterion (here: [AIC](https://en.wikipedia.org/wiki/Akaike_information_criterion)). The forward method starts with the smallest model. Within each following iteration it tests all models with one additional estimated parameter.\n",
     "\n",
     "To perform a single iteration, use `select` as shown below. Later in the notebook, `select_to_completion` is demonstrated, which performs multiple consecutive iterations automatically.\n",
     "\n",
@@ -303,7 +302,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "To search more of the model space, hence modls with more parameters, the algorithm can be repeated. As models with no estimated parameters have already been tested, subsequent `select` calls will begin with the next simplest model (in this case, models with exactly 1 estimated parameter, if they exist in the model space), and move on to more complex models.\n",
+    "To search more of the model space, hence models with more parameters, the algorithm can be repeated. As models with no estimated parameters have already been tested, subsequent `select` calls will begin with the next simplest model (in this case, models with exactly 1 estimated parameter, if they exist in the model space), and move on to more complex models.\n",
     "\n",
     "The best model from the first iteration is supplied as the predecessor (initial) model here."
    ]
@@ -443,8 +442,6 @@
     }
    ],
    "source": [
-    "from pprint import pprint\n",
-    "\n",
     "import numpy as np\n",
     "from petab_select import Model\n",
     "\n",

pypesto/visualize/select.py

@@ -21,7 +21,7 @@ def default_label_maker(model: Model) -> str:
 def plot_selected_models(
     selected_models: List[Model],
     criterion: str = Criterion.AIC,
-    relative: str = True,
+    relative: bool = True,
     fz: int = 14,
     size: Tuple[float, float] = (5, 4),
     labels: Dict[str, str] = None,