A Experiments

The likelihood-based approach to inference in this context is provided by popular packages such as MontePython [70]. Our simulator is identical to the likelihood that in MontePython is called fake_planck_realistic [71]. This likelihood is used, often in combination with other likelihoods, to forecast the expected constraining power of future experiments. In this model, the power spectra receive non-stochastic contributions from the cosmological model and the idealized measurement instrument.