Update NASA polynomial and thermdat creation functions

[wittregr]

The current polynomial fitting function in the thermochemistry code is limited. Other version of the code by Lym and Wittreich have better functionality. We need to combine the best of these codes and create a better version for the thermochemistry code. Jon Lym's code is the best to date and should be a good starting point.

[jonlym]

Referencing done using Least Squares Method (based on @wittregr's code). For fitting NASA polynomials, T_mid evaluated for each species (based on my previous code). IdealGasThermo and SimpleHarmonicThermo use ASE (based on my previous code) but it is flexible enough that we can make our own models if necessary.

Are there any other features from the old codes that need to be implemented?

[jonlym]

Based on an in-person meeting, the following action items were assigned: Gerhard

• Create Zacros class to interface Thermochemistry code with Zacros wrapper Implemented in commit a9b2b22975067a3b9c946c0f247f40443ce8f03e
• Determine if Gaussian frequencies are in 1/cm Confirmed that frequencies are in 1/cm
• Determine if Lisa has a 'slab' reference energy to subtract off Lisa gave us sample Gaussian calculation with only the slab but we need to code a way to subtract slab energy off.
• Write help file to explain conversion between old input format and new input format

Jon

• Change the excel input to take a relative reference to atoms object and add basepath to the main code. Also make 'thermo_model' field more user-friendly Implemented in commit 85a1b48db2ec5d66d430ced2766837ea56c923ec
• Add calc_offset function to the constructor of the References object Implemented in commit f6c53f5f558cb1deba6edbae6a96c501281d0eb3
• Rename Thermdat class to Nasa Implemented in commit f6c53f5f558cb1deba6edbae6a96c501281d0eb3