From bug to feature: Harnessing cross-sensitivity for multiparametric luminescence sensing.

From bug to feature: Harnessing cross-sensitivity for multiparametric luminescence sensing. This dataset supports the study titled "From bug to feature: Harnessing cross-sensitivity for multiparametric luminescence sensing", in which we employed an LDA-based (Linear Discriminant Analysis) approach to decouple the simultaneous influence of pressure and temperature on the luminescence signal of ruby (Al2O3:Cr3+). The dataset includes: Raw calibration (training and validation) spectral data: Photoluminescence spectra of a single ruby (Al2O3:Cr3+) microsphere collected at 30 pressure (P) and temperature (T) conditions for model training, and 7 P-T conditions for model validation. Each condition includes 350 spectra for robust model training and validation. Ground-truth P and T labels for the training and validation datasets. Results of the spectral data analysis concerning conventional optical readout features. Results of the LDA model training and validation. The data are organized into four sections (a total of 51 files): 1-37) raw training and validation spectral data, 38) Conventional optical readout features, 39-42) LDA training and validation, and 43-51) LDA results. All data are stored in .csv format, for the exception of file 38 (.xlxs). These data support key findings presented in the manuscript and enable verification of the (indicated) figures. Section 1: Raw training and validation spectral data File: 1. 298.9_0.00.csv File: 2. 298.6_2.44.csv File: 3. 298.5_2.61.csv File: 4. 298.5_4.47.csv File: 5. 298.6_4.49.csv File: 6. 323.5_0.00.csv File: 7. 323.4_2.51.csv File: 8. 323.4_3.92.csv File: 9. 323.4_4.97.csv File: 10. 347.8_0.00.csv File: 11. 347.6_2.52.csv File: 12. 347.6_3.58.csv File: 13. 347.6_4.76.csv File: 14. 374.0_0.00.csv File: 15. 374.0_2.21.csv File: 16. 374.1_2.53.csv File: 17. 374.0_4.61.csv File: 18. 398.7_0.00.csv File: 19. 398.7_1.43.csv File: 20. 398.8_2.54.csv File: 21. 398.9_3.75.csv File: 22. 398.7_4.34.csv File: 23. 424.0_0.00.csv File: 24. 424.1_0.75.csv File: 25. 424.0_1.25.csv File: 26. 424.2_2.12.csv File: 27. 424.1_2.25.csv File: 28. 424.3_2.42.csv File: 29. 424.2_3.93.csv File: 30. 424.3_4.07.csv File: 31. 312.8_0.00_Validation.csv File: 32. 385.8_0.00_Validation.csv File: 33. 413.7_1.67_Validation.csv File: 34. 362.3_2.42_Validation.csv File: 35. 407.7_2.44_Validation.csv File: 36. 334.0_4.11_Validation.csv File: 37. 389.8_4.23_Validation.csv Description: Raw spectral data, organized by condition. Used for building Figures 1-3 of the manuscript. The first column within each file contains the wavelength axis, and columns 2-351 contain the 350 spectra. The corresponding ground-truth T and P conditions at which the spectra were acquired are indicated by the file name, e.g., 298.9_0.00 corresponds to 298.9 K and 0.00 GPa. Section 2: Conventional optical readout features File: 38. Conventional optical readout features.xlsx Description: Summary of all results concerning the conventional optical readout features. Used to build Figures 2 and 5b of the manuscript. Section 3: LDA training and validation File: 39. Training Data_30 Conditions File: 40. Validation Data_7 conditions File: 41. T and P Ground Truth Labels_Training Data File: 42. T and P Ground Truth Labels _Validation Data Description: Same dataset as in Section 1, but consolidated into fewer .csv files for LDA training and validation. The .csv file structure is compatible with the code file (LDA for harnessing cross-sensitivity_model training and validation_results data export.ipynb), which can be accessed following the DOI link provided in the “Code Availability” section of the manuscript. Section 4: LDA results File: 43. LD1 P_explained variance ratio File: 44. LD1 T_explained variance ratio File: 45. LD1 P_weighting coefficients File: 46. LD1 T_weighting coefficients File: 47. LD1 P_mean scores vs P and T_training conditions File: 48. LD1 T_mean scores vs P and T_training conditions File: 49. LD1 P_predicted vs true P and residuals_linear model File: 50. LD1 T_predicted vs true T and residuals_cubic model File: 51. LD1 T and LD1 P_mean scores vs P and T_validation conditions Description: Files containing the main LDA results, exported from the aforementioned code file (.ipynb) developed using Python in the Jupyter Notebook environment. Used to build Figures 4 and 5a of the manuscript. All other LDA-related figures presented in the Supporting Information can be exported from the same .ipynb file.