respy.config

General configuration for respy.

Module Contents

respy.config.CHAOSPY_INSTALLED = False

respy.config.ROOT_DIR

respy.config.TEST_DIR

respy.config.TEST_RESOURCES_DIR

respy.config.MAX_FLOAT = 1e+200

respy.config.MIN_FLOAT

respy.config.MAX_LOG_FLOAT = 460

respy.config.MIN_LOG_FLOAT

respy.config.COVARIATES_DOT_PRODUCT_DTYPE

numpy.dtype : Dtype of covariates before being used in a dot product.

If you convert a DataFrame with boolean variables and others to an NumPy array, the resulting array will have an ‘object’ dtype. Having an ‘object’ dtype array causes a lot of problems as functions like numpy.exp() will fail raising an uninformative error message.

respy.config.DTYPE_STATES

respy.config.INDEXER_DTYPE

numpy.dtype : Data type for the entries in the state space indexer.

respy.config.INDEXER_INVALID_INDEX

int : Identifier for invalid states.

Every valid state has a unique number which is stored in the state space indexer at the correct position. Invalid entries in the indexer are filled with INDEXER_INVALID_INDEX which is the most negative value for INDEXER_DTYPE. Using the invalid value as an index likely raises an IndexError as negative indices cannot exceed the length of the indexed array dimension.

respy.config.TOL_REGRESSION_TESTS = 1e-10

respy.config.SEED_STARTUP_ITERATION_GAP = 1000000

respy.config.DEFAULT_OPTIONS

respy.config.KEANE_WOLPIN_1994_MODELS

respy.config.KEANE_WOLPIN_1997_MODELS = ['kw_97_basic', 'kw_97_basic_respy', 'kw_97_extended', 'kw_97_extended_respy']

respy.config.KEANE_WOLPIN_2000_MODELS = ['kw_2000']

respy.config.ROBINSON_CRUSOE_MODELS = ['robinson_crusoe_basic', 'robinson_crusoe_extended', 'robinson_crusoe_with_observed_characteristics']

respy.config.EXAMPLE_MODELS