Unveiling the origin of fast radio bursts by modeling the stellar mass and star formation distributions of their host galaxies

Diverse formation channels have been proposed to explain the emergence of fast radio bursts (FRBs), yet their origin remains elusive. With improved localization precision, roughly 90 FRBs are now associated with host galaxies. SED fitting to the host galaxy photometry reveals their stellar masses ($M_\star$) and star formation rates (SFRs), enabling discrimination between various formation channels. We conduct an extensive comparison of the stellar mass, SFR and z distributions of 51 FRB hosts and mock-generated galaxy samples to test whether FRBs trace SFR or $M_\star$. We incorporate a mass-to-light ratio prescription to address optical selection biases. In line with Sharma et al. (2024), we provide evidence in favor of FRB progenitors tracking SF rather than stellar mass. We show that the shape of the assumed $(M_\star/L_r)_{obs}$ distribution affects the predictions, bringing the low mass end of the stellar mass distribution closer to the data when accounting for the $\mathrm{SFR} - (M_\star/L_r)_{obs}$ correlation. The K-correction effect in the $r-$band is minimal for galaxies at $z \lesssim 0.7$. Even if FRBs trace SF, up to $\sim 6\%$ of a flux-limited FRB host sample can reside below the star-forming main sequence. Finally, we examine a hybrid model in which a fraction of FRBs tracks stellar mass rather than SF. This fraction can be as large as $\sim(40-50)\%$, suggesting that multiple formation channels are still consistent with observations. The toolkit developed in this work is publicly available (GALFRB code), offering a straightforward way to generate mock galaxy samples suitable for direct comparisons with future FRB host galaxy data.