[TASK] Expose measurement likelihoods from PDAF for IPDA integration

[jorgenfj]

Description of task

Currently, the IPDA filter recomputes measurement likelihoods (z_pred.pdf(z_k)) inside existence_prob_update().
However, these same likelihoods are already calculated internally in the PDAF during association weighting (get_weights()).

This causes duplicated computation and makes it harder to integrate IPDA and PDAF efficiently.

---

Proposed change

Extend the PDAF::Output struct and PDAF::step() to include a vector of measurement likelihoods for the gated measurements.

Example:

struct Output {
    Gauss_x x_;
    Gauss_x x_prediction;
    Gauss_z z_prediction;
    Gauss_xX x_updates;
    Arr_1Xb gated_measurements;
    Eigen::ArrayXd measurement_likelihoods;  // new field
};

In step(), compute and populate:

Eigen::ArrayXd z_likelihoods(inside_meas.cols());
for (int i = 0; i < inside_meas.cols(); ++i)
    z_likelihoods(i) = z_pred.pdf(inside_meas.col(i));

and include it in the returned output.

---

Benefits

• Eliminates redundant likelihood computation inside IPDA.
• Keeps likelihood generation within the PDAF, where it naturally belongs.
• Simplifies the IPDA implementation (can use the existence_prob_update(likelihoods, ...) overload).
• Improves performance and clarity.

---

Example integration

After this change, IPDA::step() could do:

auto [x_post, x_pred, z_pred, x_upd, gated_measurements, z_likelihoods] =
    PDAF::step(...);

double existence_probability_upd =
    IPDA::existence_prob_update(z_likelihoods, existence_prob_pred, config);

---

Suggested Workflow

No response

Specifications

No response

Contacts

No response

Code Quality

• Every function in header files are documented (inputs/returns/exceptions)
• The project has automated tests that cover MOST of the functions and branches in functions (pytest/gtest)
• The code is documented on the wiki (provide link)