Problem
Let's say we are learning a classifier for the `Iris` dataset.
```python
from sklearn.datasets import load_iris
X, y_true = load_iris(return_X_y=True, as_frame=True)
```
The classifier is a `RandomForestClassifier` learned as follows.
```python
from sklearn.ensemble import RandomForestClassifier
model = RandomForestClassifier(n_estimators=2)
model.fit(X, y_true)
```
Compute the top-k accuracy measure for k=1 and k=2.