We apply the massively optimized parameter estimation and data compression technique (MOPED) to the public Planck 2015 temperature likelihood, reducing the dimensions of the data space to one number per parameter of interest. We present CosMOPED, a lightweight and convenient compressed likelihood code implemented in Python. In doing so we show that the $\ell<30$ Planck temperature likelihood can be well approximated by two Gaussian distributed data points, which allows us to replace the map-based low-$\ell$ temperature likelihood by a simple Gaussian likelihood. We make available a Python implementation of Planck's 2015 Plik_lite temperature likelihood that includes these low-$\ell$ binned temperature data (Planck-lite-py). We do not explicitly use the large-scale polarization data in CosMOPED, instead imposing a prior on the optical depth to reionization derived from these data. We show that the $\Lambda$CDM parameters recovered with CosMOPED are consistent with the uncompressed likelihood to within 0.1$\sigma$, and test that a 7-parameter extended model performs similarly well.

We report on measurements of the polarization of extragalactic sources at 148 GHz made during the first two seasons of the Atacama Cosmology Telescope Polarization (ACTPol) survey. The survey covered 680 deg$^{2}$ of the sky on the celestial equator. Polarization measurements of 169 intensity-selected sources brighter than 30 mJy, that are predominantly Active Galactic Nuclei, are presented. Above a total flux of 215 mJy where the noise bias removal in the polarization measurement is reliable, we detect 26 sources, 14 of which have a detection of linear polarization at greater than 3$\sigma_{p}$ significance. The distribution of the fractional polarization as a function of total source intensity is analyzed. Our result is consistent with the scenario that the fractional polarization of our measured radio source population is independent of total intensity down to the limits of our measurements and well described by a Gaussian distribution with a mean fractional polarization $p=0.028\pm$0.005 and standard deviation $\sigma_{\mathrm{p}}=0.054$, truncated at $p=0$. Extrapolating this model for the distribution of source polarization below the ACTPol detection threshold, we predict that one could get a clean measure of the E-mode polarization power spectrum of the microwave background out to $\ell=6000$ with 1 $\mu$K-arcminute maps over 10$\%$ of the sky from a future survey. We also study the spectral energy distribution of the total and polarized source flux densities by cross-matching with low radio frequency catalogs. We do not find any correlation between the spectral indices for total flux and polarized flux.

We develop a method of reconstructing the lensing field from lensed CMB temperature and polarization maps in real space as an alternative to the harmonic space estimators currently in use by extending an existing real space lensing estimator for temperature to polarization. Real space estimators have the advantage of being local in nature and they are thus equipped to deal with the nonuniform sky coverage, especially galactic cuts and point source excisions, found in experimental data. We characterize some of the properties and limitations of these estimators and test them on simulated maps with Planck, AdvACT and CMB-S4 noise. We show that the reconstructions for large-scale lensing fields are accurate, and that the polarization reconstructions improve on those from CMB temperature maps for future experiments as expected. High-fidelity lensing maps can be reconstructed with futuristic experiments like CMB-S4.