READ ME file for 'data.zip' Dataset DOI: 10.5258/SOTON/D2104 ReadMe Author: S. L. Harrison, Hybrid Photonics Group, University of Southampton, S.L.Harrison@soton.ac.uk This dataset supports the publication: AUTHORS: Stella L. Harrison, Helgi Sigurdsson, Sergey Alyatkin, Julian D. Töpfer, Pavlos G. Lagoudakis TITLE: Solving the max-3-cut problem with coherent networks JOURNAL: American Physical Society: Physical Review Applied PAPER DOI IF KNOWN: TBC ____________________________________________________________________________________________________________ This dataset contains: ** fig_1d.csv ********************************************************************************************** Fig 1(d) - size: 4 x 5 Histrogram of max-3-cut weights from house graph Row 1: Histogram bar weight of cut (2 to 6 in integer steps) Row 2: Stochastic weight of cut bar height Row 3: Simulated weight of cut bar height Row 4: Experimental weight of cut bar height + vertical line @ weight = 6 ** fig_1e.csv ********************************************************************************************** Fig 1(e) - size: 4 x 7 Histogram of max-3-cut energies from house graph Row 1: Histogram bar energies Row 2: Stochastic energy bar height Row 3: Simulated energy bar height Row 4: Experimental enery bar height + vertical line @ energy = -8.7419 ** fig_1f.csv ********************************************************************************************** Fig 1(f) - size: 1200 x 1200 Real space image of house excitation profile (2DGPE simulation) 1 pixel = 1/15 µm (-40 µm to 40 µm over 1200 pixels, in x and y directions) ** fig_1g.csv ********************************************************************************************** Fig 1(g) - size: 240 x 240 Real space photoluminescence from house excitation profile (2DGPE simulation) 1 pixel = 0.4 µm (-48 µm to 48 µm over 240 pixels, in x and y directions) ** fig_1h.csv ********************************************************************************************** Fig 1(h) - size: 240 x 240 Phase space of photoluminescence from house excitation profile (2DGPE simulation) 1 pixel = 0.4 µm (-48 µm to 48 µm over 240 pixels, in x and y directions) ** fig_1i.csv ********************************************************************************************** Fig 1(i) - size: 460 x 460 Real space image of house excitation profile (experiment) - time integrated 1 pixel = 0.1296 µm (in x and y directions) ** fig_1j.csv ********************************************************************************************** Fig 1(j) - size: 927 x 926 Real space photoluminescence from house excitation profile (experiment) 1 pixel = 0.0648 µm (in x and y directions) ** fig_1k.csv ********************************************************************************************** Fig 1(k) - size: 927 x 927 Phase space of photoluminescence from house excitation profile (experiment) 1 pixel = 0.0648 µm (in x and y directions) ** fig_4a.csv ********************************************************************************************** Fig 4(a) - size: 298 x 864 Apple image RGB values levels in 298 x 288 images concatenated horrizontally ** fig_4g.csv ********************************************************************************************** Fig 4(g) - size: 256 x 768 Tree image RGB values levels in 256 x 256 images concatenated horrizontally ** fig_5a.csv ********************************************************************************************** Fig 5(a,b) - size: 5 x 20 Simple image + weights matrix RGB values levels in 5 x 5 images concatenated horrizontally + 5 x 5 weights matrix ** fig_5c.csv ********************************************************************************************** Fig 5(c) - size: 91 x 91 Complex wavefunction (amplitude and phase) of simple image segmentation (2DGPE) 1 pixel = 0.5 µm (in x and y directions) ** fig_6f.csv ********************************************************************************************** Fig 6(f) - size: 161 x 161 Complex wavefunction (amplitude and phase) of CVM (2DGPE) 1 pixel = 0.4 µm (in x and y directions) ** fig_7a.csv ********************************************************************************************** Fig 7(a) - size: 926 x 926 Real space photoluminescence from square excitation profile (experiment) 1 pixel = 0.0648 µm (in x and y directions) ** fig_7b.csv ********************************************************************************************** Fig 7(b) - size: 468 x 468 Reciprocal (momemtum) space photoluminescence from square excitation profile (experiment) 1 pixel = 0.00855 µm^-1 (in x and y directions) ** fig_7c.csv ********************************************************************************************** Fig 7(c) - size: 927 x 927 Real space photoluminescence from house excitation profile (experiment) - single shot 1 pixel = 0.0648 µm (in x and y directions) ** fig_8a.csv ********************************************************************************************** Fig 8(a) - size: 2 x 160 weight of cut from max-2-cut Row 1: weight of cuts solved using Stuart-Landau model Row 2: weight of cuts solved using brute force method ** fig_8b.csv ********************************************************************************************** Fig 8(b) - size: 2 x 160 weight of cut from max-3-cut Row 1: weight of cuts solved using Stuart-Landau model Row 2: weight of cuts solved using brute force method ** fig_8c.csv ********************************************************************************************** Fig 8(c) - size: 2 x 160 weight of cut from max-4-cut Row 1: weight of cuts solved using Stuart-Landau model Row 2: weight of cuts solved using brute force method ** fig_8d.csv ********************************************************************************************** Fig 8(d) - size: 4 x 9 Average error of Stuart-Landau weight of cut vs. number of vertices Row 1: Number of vertices (6 to 14 in integer steps) Row 2: Average max-2-cut error Row 3: Average max-3-cut error Row 4: Average max-4-cut error ** fig_8e.csv ********************************************************************************************** Fig 8(e) - size: 4 x 11 Average error of Stuart-Landau weight of cut vs. number of binning boundaries tested Row 1: Number of binning boundaries Row 2: Average max-2-cut error Row 3: Average max-3-cut error Row 4: Average max-4-cut error