[Access this dataset on Dryad] https://doi.org/10.5061/dryad.xwdbrv1mw

Description of the data and file structure

1. 112323_A1_C3R3_Drop_Group_Delay_larger_range.txt

Contains linear transmission and group delay data from Fig. 2 of the manuscript. Format is simple .txt file. Variables, which are columns in the text file, are labeled. In order, they are: Wavelength (nm), Frequency (GHz), Insertion Loss (dB), Group Delay (ps), Chromatic Dispersion (ps/nm).
2. 231122_Run_3_Chip_A1_Col_3_Row_3_HighResOSA_Drop_bkTel_Theta_285__bkTel_swl_1547p35_OSA_N_1_centerwav_1500p0_span_400p0.npz

231122_Run_3_Chip_A1_Col_3_Row_3_HighResOSA_Drop_bkTel_Theta_283__bkTel_swl_1547p35_OSA_N_1_centerwav_1500p0_span_400p0.npz

231122_Run_3_Chip_A1_Col_3_Row_3_HighResOSA_Drop_bkTel_Theta_280__bkTel_swl_1547p35_OSA_N_1_centerwav_1500p0_span_400p0.npz

231122_Run_3_Chip_A1_Col_3_Row_3_HighResOSA_Drop_bkTel_Theta_278__bkTel_swl_1547p35_OSA_N_1_centerwav_1500p0_span_400p0.npz

231122_Run_3_Chip_A1_Col_3_Row_3_HighResOSA_Drop_bkTel_Theta_275__bkTel_swl_1547p35_OSA_N_1_centerwav_1500p0_span_400p0.npz

Contain Optical Spectrum Analyzer (OSA) spectra for five different pump powers as shown in Fig. 3 of the manuscript. (275, 278, 280, 283, 285 correspond to {70, 78, 86, 92, 100} W of on-chip peak power.) These can be processed via simple python function shown below. Variables stored are "intensity", which is a numpy array of intensity values collected by the OSA, "times" which is a numpy array which stores timestamps for each datapoint in the "intensity" (dB) array, and "wl", a numpy array storing wavelength (nm) values that correspond to each value in "intensity". 
3. 240311_Run_3_A1_C3R3_Thru_Sample_Theta_24__FSV_BW_1kHz_AttnESA_0db_pts_31001_0_to_5p0GHz_wAPEX_bkTel_swl_1547p35_OSA_N_3_centerwav_1548p0_span_500p0.npz

240311_A1_C3R3_Thru_ApexAttn0p005_Theta_24_1547p35_Spectrum.dat

These contain data shown in Fig. 4. The .npz file contains the low resolution OSA spectrum stored as described above in #2. High resolution nested data from the ultra-high resolution heterodyne-based optical spectrum analyzer is stored in the .dat file. The .dat processing script below returns two variables: freq_in_GHz gives the x-axis of the collected spectrum in GHz, signal_in_vol gives the y-axis of the collected spectrum in voltage. 
4. CCW_v1.png

CCW_Bulk_v1.png

CW_v1.png

These are the raw images taken that were then processed for Fig. 5 in the manuscript. CCW_v1 corresponds to panel A, CW_v1 to panel B, and CCW_Bulk_v1 to panel C. 

Python Code For Processing

1. npz files can be processed with simple python function as follows.

   import numpy as np
   def load_sweep_file(filename): 
       npzfile = np.load(filename)
       intensity = npzfile['intensity']
       times = npzfile['times']
       wl = npzfile['wl']
       return intensity, times, wl

1. .dat file requires a more complex script, but can be processed with the following:

   # Created by Apex Technologies
   import os
   def read_spectrum_binary_file(path, name_file):
       '''
       This function is to read spectrum binary file saved from OSA.
       The content of the spectrum binary stream is
          b185000.000000 7876 kh\x02Fk...
          b is a character of 'b' written into the binary file.
          185000.000000 is defaul optical frequency base.
          7876 is number of points.
          kh\x02Fk... stream contains the encoded frequency and signal voltage.
       Note: In this binary stream, the encoded optical frequency is in GHz,
             signal is in Voltage.
       '''

       #  Check if existing the path.
       if os.path.exists(path) == False:
           try:
               os.mkdir(path)
           except OSError:
               print("Creation of the directory %s failed" % path + '\n')
           else:
               print("Successfully created the directory %s " % path + '\n')
       else:
           print("the directory exits %s " % path + '\n')

       #  Open and reading the file to get the data.
       with open(path + '/' + name_file, 'rb') as binary_file:

           # Read the first character 'b' in the binary sream.
           ch = binary_file.read(1).decode('utf-8')

           if ch != 'b':
               path_file = path + '/' + name_file
               print("The file %s" % path_file + "is not binary file.")
               return

           # Read the default optical frequency base.
           optical_freq_base_GHz = binary_file.read(14).decode('utf-8')

           # Read the number of points.
           number_of_points = ''
           while True:
               ch = binary_file.read(1).decode('utf-8')
               number_of_points = number_of_points + ch
               if ch == str(' '):
                   break

           number_of_points = int(number_of_points)
           # Read and unpack the double-precision float frequency in GHz.
           format_string = f'{number_of_points}f'
           freq_in_GHz = struct.unpack(
               format_string, binary_file.read(struct.calcsize(format_string)))

           # Read and unpack the double-precision float signal in Voltage.
           signal_in_vol = struct.unpack(
               format_string, binary_file.read(struct.calcsize(format_string)))

           # Convert tuple to list type.
           freq_in_GHz = np.array(freq_in_GHz)
           signal_in_vol = np.array(signal_in_vol)

           # Decode frequency vector to real one.
           freq_in_GHz += float(optical_freq_base_GHz)

           return freq_in_GHz, signal_in_vol