Consistent data format for analysis results

[marcolincs]

Current behavior

Output data is stored in self._variables in QAnalysis (accessible with self.get_data()). However the format varies for each return element, forcing the user to interpret the data as needed.

Requested behavior

Create a uniform data structure. Current ideas are:

• flat data expression. indexes, unit measures and numerical value all stored in separate objects within a dictionary
• pandas with multiple indexes. Everything is described inside the pandas dataframe, except for the units, therefore they need to be stored as two separate objects within a dictionary

Below two examples (not completely defined):

resultnp = Dict(
    cap_matrix = Dict(
        index1 = [200,220],
        index2 = [1,2],
        index3 = ["Cin","Cout"],
        index4 = ["Cin","Cout"],
        values = a,
        units = ["nF","um",       None,    None,  None  ],
        indexes =    ['pad_width','passes','CapY','CapX']
    ))

resultpd_i = Dict(
    cap_matrix = Dict(
        values = index_select_df,
        units = ["nF","um",None,None,None]
    ))

where

b = a.reshape(16)
index_select_df = pd.DataFrame(zip(c,d,e,f,b),columns=['CapX','CapY','passes','pad_width','CapValue'])
index_select_df.set_index(['CapX','CapY','passes','pad_width'])