Restructuring variable transformations

• posterior_utils.py now only has sampling algorithms
• all variable transformations are done using functions in transforms.py
• posterior_utils.generate_obs_from_true_list now only outputs samples in chirp mass, symmetric mass ratio, SNR, and theta
• transforms.transform converts these into the parameters that the user wants
• transforms.transform computes jacobians for these transforms and reports the induced prior in terms of whatever parameters were asked for
• generate-mock-PE.py contains an pseudo-config file as the first several lines of code where the user can specify which variables and prior they want. There are limited options, but they should cover anything wanted in a hierarchical population inference.