Source data for publication 'Modulatability of complex oscillators'

This package contains the numerical source data underlying the data-bearing main, Extended Data and Supplementary figures in the associated article. Contents `data/`: source data grouped by scientific case. `SOURCE_DATA_INDEX.csv`: panel-level mapping from figure panels to data files. `DATA_DICTIONARY.md`: definitions of shared fields, units and missing values. `SHA256SUMS.csv`: byte counts and SHA-256 checksums for all files in `data/`. `LICENSE`: reuse terms for this dataset. File formats CSV files use UTF-8 text, a single header row and comma-separated values. Electronic-circuit measurements are supplied as original tab-delimited TXT files and UTF-8 normalized CSV copies. `SOURCE_DATA_INDEX.csv` is the authoritative mapping from the current manuscript panels to package files; file names and column headers identify the scientific case, condition and physical quantity. Indices are one-based. Time and phase units are stated in column headers, if applicable. `NaN` denotes an unavailable or inapplicable numerical value; an empty text field denotes unavailable text metadata. Additional shared definitions are in`DATA_DICTIONARY.md`. Values stored in provenance columns such as `source_file` and `result_file` identify the repository-side intermediate from which a row was exported. They are provenance labels rather than paths that must resolve inside this archive. Figure-specific scope Figure mappings in `SOURCE_DATA_INDEX.csv` are ordered first by figure class (Main Figures, Extended Data Figures, Supplementary Figures), then by figurenumber, panel letter and subpanel numeral. Each ordinary row identifies exactly one panel. When a panel requires multiple files, all required relative paths arelisted in the same `file` cell and separated by `; `. A data file may therefore appear in more than one row when it supports more than one panel. Additional network robustness files that are not directly plotted are collected in the single `Supporting network dataset (not directly plotted)` entry. Main Fig. 1 uses the GRN model on a Barabasi-Albert network with 100 nodes. Panels c(i) and c(ii) show the baseline and modulated network structures, respectively. Panels d(i) and d(ii) show the corresponding baseline and modulated oscillation phase trajectories; the modulated condition uses `n_per=50`. Panel e shows node amplitudes and panel f shows their distributions. Panels g and h use the ergodic dataset with target-node `sequence=1` and modulation `threshold=0.05`. All panels in Main Fig. 2 and Main Fig. 3e are schematic and therefore have no source-data entry. The data-bearing panels of Main Fig. 3 are b, c, d and f. Extended Data Fig. 1 uses the corresponding FHN/BA/N=100 datasets. The network Supplementary figures use the following datasets: S1--S4: representative and ergodic results for GRN/ER, GRN/SW, FHN/ER and FHN/SW, respectively (`N=100`, `sequence=1`, `threshold=0.05`); each representative modulated condition has 50 perturbed edges. S5--S6: GRN and FHN BA network-size sweeps for `N=50,100,200`; the scatter slices use `n_target=25,50,100`, respectively, with `sequence=1` and `threshold=0.05`. S7--S8: GRN and FHN BA target-selection sweeps for `sequence=1,2,3` (`N=100`, `threshold=0.05`). S9--S10: GRN and FHN BA threshold sweeps for threshold=0.05,0.075,0.1` (`N=100`, `sequence=1`). S11--S12: GRN and FHN exclusion rates, respectively. All packaged data files are listed in `SOURCE_DATA_INDEX.csv`. Integrity and citation `SHA256SUMS.csv` can be used to verify the integrity of every data file. Please cite the associated article and the DOI assigned to this dataset record whenreusing the data. This dataset is licensed under the Creative Commons Attribution 4.0 International License; see `LICENSE`.