Artificial dataset generated to train C4TUNE
收藏资源简介：
This dataset was created to train a neural network to predict the parameters of a kinetic model of C<sub>4</sub> photosynthesis (Wang et al. 2021, doi: 10.1111/tpj.15365). In essence, parameters of the kinetic models were sampled from a multivariate log-normal distribution, and for each sample, response curves of the net CO<sub>2</sub> assimilation rate, A<sub>net</sub>, to varying ambient CO<sub>2</sub> partial pressures (Ca) and light intensities (Q) were simulated using the ODE model. The simulated curves and associated parameterizations were filtered by technical and biological criteria as described in the publication. This dataset contains the following files: File name Description params.csv parameter samples a_co2.csv simulated A/Ca curves a_light.csv simulated A/Q curves co2_a_light.txt constant Ca for A/Q curves light_a_co2.txt constant Q for A/Ca curves train_idx_c4tune.npy indices of the training set for C4TUNE (numpy) train_idx_c4tune.txt indices of the training set for C4TUNE (text file) train_idx_surrogate.npy indices of the training set for the surrogate model (numpy) train_idx_surrogate.txt indices of the training set for the surrogate model (text file) test_idx_c4tune.npy indices of the test set for C4TUNE (numpy) test_idx_c4tune.txt indices of the test set for C4TUNE (text file) test_idx_surrogate.npy indices of the test set for the surrogate model (numpy) test_idx_surrogate.txt indices of the training set for the surrogate model (text file) L_train.csv Cholesky decomposition of the parameter covariance matrix, based on the training set L_test.csv Cholesky decomposition of the parameter covariance matrix, based on the test set
