According to the current paradigm of cosmological structure formation,
the observable large scale matter distribution arose via gravitational
amplification from tiny primordial density fluctuations. Especially
modern numerical simulations reveal that cold dark matter aggregates to
form a filamentary cosmic web consisting of huge empty regions, the so
called voids, filaments and clusters. Hence, precision analysis of three
dimensional large scale structure (LSS) data will help us to identify
and understand the physical processes governing cosmological structure
formation leading to a more complete theoretical picture of our Universe.

However, contact between theory and observations cannot be made
directly, since observational data is subject to a variety of systematic
effects and statistical uncertainties. Most notably of those are the
survey geometry and selection effects as well as statistical noise.
Mapping the three dimensional matter distribution in the Universe thus
requires accurate statistical data analysis methods. In my talk I will
present new full Bayesian data analysis methods designed to provide
detailed cosmographic descriptions of the large scale structure in the
Universe while accounting for all observational uncertainties. As a
result these methods provide sampled representation of the LSS posterior
distribution, which enables us to report any desired statistical summary
such as mean, mode or variance of the density field. Application of our
method to the latest Sloan Digital Sky Survey data
lead to the generation of detailed cosmographic maps for the three
dimensional matter distribution and the possibility to accurately
quantify its significance. These results permit a variety of following
scientific projects to analyze the clustering behavior of matter in the
Universe. In summary, the presented methods provide an efficient and
flexible basis for high-precision LSS inference.