Data from fig 3 is the same dataset as Fig. 1.

For 3(a) we calculate process the data to get the fractional change in absorption ((on-off)/(off)) for the microwave frequency 15.0 GHz file.

For 3(b) we construct the heat map in fig. 1 and take cross section of this at particular energies (energies used are detailed in manuscript) We then take the ratio of these cross sections to get teh data plotted in 3(b).

for the model we use the parameters in the files: 

These paramters come from fitting the spectra (with no microwave field) in figs 1 and 2. The sprecum in fig. 1 is fitted with a sum of asymmetric Lorentzians (see Nature volume 514, pages343–347 (2014)) and the spectrum in fig 2 is fitted with a sum of symmetric lorentzians. Definition of form of lorentzian is given in fit_para file.

The dipole matrix elements used were theoretically calculated. These are tubulated in the sub folder "DME". each file is named as a state and the matrix elements are transitions from this state to the other states labelled in the file. The units are um.



The model itself was written in Mathematica, the code is in this folder.

 