BUG: Provide missing position arg to local cross validation function

Provide missing position argument to local cross validation function in
GP estimation error analysis script.

Fixes
```
scripts/dwi_gp_estimation_error_analysis.py:210: error:
 Missing positional argument "gpr" in call to "cross_validate"  [call-arg]
scripts/dwi_gp_estimation_error_analysis.py:210: error:
 Unsupported operand types for * ("float" and "dict[int, list[tuple[ndarray[Any, Any], ndarray[Any, Any], ndarray[Any, Any], ndarray[Any, Any]]]]")  [operator]
scripts/dwi_gp_estimation_error_analysis.py:210: error:
 Argument 4 to "cross_validate" has incompatible type "DiffusionGPR"; expected "int"  [arg-type]
scripts/dwi_gp_estimation_error_analysis.py:211: error:
 "float" has no attribute "tolist"  [attr-defined]
scripts/dwi_gp_estimation_error_analysis.py:212: error:
 Argument 1 to "len" has incompatible type "float"; expected "Sized"  [arg-type]
scripts/dwi_gp_estimation_error_analysis.py:213: error:
 Argument 1 to "len" has incompatible type "float"; expected "Sized"  [arg-type]
scripts/dwi_gp_estimation_error_analysis.py:214: error:
 Argument 1 to "len" has incompatible type "float"; expected "Sized"  [arg-type]
scripts/dwi_gp_estimation_error_analysis.py:215: error:
 Argument 1 to "len" has incompatible type "float"; expected "Sized"  [arg-type]
```

raised for example in:
https://github.com/nipreps/nifreeze/actions/runs/12437972140/job/34728973936#step:8:113

scripts/dwi_gp_estimation_error_analysis.py

@@ -202,12 +202,14 @@ def main() -> None:
         # max_iter=2e5,
     )
 
+    n_repeats = 10
+
     if args.kfold:
         # Use Scikit-learn cross validation
         scores = defaultdict(list, {})
         for n in args.kfold:
             for i in range(args.repeats):
-                cv_scores = -1.0 * cross_validate(X, y.T, n, gpr)
+                cv_scores = -1.0 * cross_validate(X, y.T, n, n_repeats, gpr)
                 scores["rmse"] += cv_scores.tolist()
                 scores["repeat"] += [i] * len(cv_scores)
                 scores["n_folds"] += [n] * len(cv_scores)