rail.estimation.algos.sklearn_nn module

Example code that implements a simple Neural Net predictor for z_mode, and Gaussian centered at z_mode with base_width read in fromfile and pdf width set to base_width*(1+zmode).

class rail.estimation.algos.sklearn_nn.Inform_SimpleNN(args, comm=None)

Bases: CatInformer

Subclass to train a simple point estimate Neural Net photoz rather than actually predict PDF, for now just predict point zb and then put an error of width*(1+zb). We’ll do a “real” NN photo-z later.

config_options = {'bands': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}, 'hdf5_groupname': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}, 'mag_limits': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}, 'max_iter': <ceci.config.StageParameter object>, 'nondetect_val': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}, 'nzbins': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}, 'output_mode': <ceci.config.StageParameter object>, 'redshift_col': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}, 'ref_band': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}, 'save_train': True, 'width': <ceci.config.StageParameter object>, 'zmax': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}, 'zmin': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}}

name = 'Inform_SimpleNN'

run()

Train the NN model

class rail.estimation.algos.sklearn_nn.SimpleNN(args, comm=None)

Bases: CatEstimator

Subclass to implement a simple point estimate Neural Net photoz rather than actually predict PDF, for now just predict point zb and then put an error of width*(1+zb). We’ll do a “real” NN photo-z later.

config_options = {'bands': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}, 'chunk_size': 10000, 'hdf5_groupname': <class 'str'>, 'nondetect_val': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}, 'output_mode': <ceci.config.StageParameter object>, 'ref_band': {'bands': ['mag_u_lsst', 'mag_g_lsst', 'mag_r_lsst', 'mag_i_lsst', 'mag_z_lsst', 'mag_y_lsst'], 'dz': 0.01, 'err_bands': ['mag_err_u_lsst', 'mag_err_g_lsst', 'mag_err_r_lsst', 'mag_err_i_lsst', 'mag_err_z_lsst', 'mag_err_y_lsst'], 'hdf5_groupname': 'photometry', 'mag_limits': {'mag_g_lsst': 29.04, 'mag_i_lsst': 28.62, 'mag_r_lsst': 29.06, 'mag_u_lsst': 27.79, 'mag_y_lsst': 27.05, 'mag_z_lsst': 27.98}, 'nondetect_val': 99.0, 'nzbins': 301, 'redshift_col': 'redshift', 'ref_band': 'mag_i_lsst', 'zmax': 3.0, 'zmin': 0.0}, 'width': <ceci.config.StageParameter object>}

name = 'SimpleNN'

rail.estimation.algos.sklearn_nn.make_color_data(data_dict, bands, ref_band, nondet_val)

make a dataset consisting of the i-band mag and the five colors

Returns:

input_data

Return type:

ndarray array of imag and 5 colors

rail.estimation.algos.sklearn_nn.regularize_data(data)

Utility function to prepare data for sklearn