csv.cov_same_xam¶

View page source

Example and Test of covariate_same¶

Covariates¶

This example has two covariates, haqi and income.

haqi¶

This covariate only depends on age and time; i.e., it is the same for all nodes and both sexes.

income¶

This covariate depends on age, time, and sex. It is the same for all nodes.

covariate_table¶

The covariate table is shuffled, after it is created, to demonstrate the fact that the order of its rows does not matter.

cov_same¶

cov_same is checked to see that it detects the following:

haqi¶

This covariate is the same for all nodes and both sexes.

income¶

This covariate is different for each sex, but given sex it is the same for all nodes.

Source¶

# at_cascade
# import at_cascade with a preference current directory version
import os
import sys
import random
current_directory = os.getcwd()
if os.path.isfile( current_directory + '/at_cascade/__init__.py' ) :
    sys.path.insert(0, current_directory)
import at_cascade
#
# haqi
def haqi(node, sex, age, time) :
    value = age / 100.0 + (time - 2000) / 20.0 + 0.01
    value = round(value, 4)
    return value
#
# income
def income(node, sex, age, time) :
    value = age / 100.0 + (2000 - time) / 20.0
    if sex == 'female' :
        value += 1.0
    else :
        value += 2.0
    value = round(value, 4)
    return value
#
# main
def main() :
    #
    age_list  = [ 20.0, 80.0 ]
    time_list = [ 1980.0, 2020.0 ]
    node_list = [ 'n0', 'n1', 'n2' ]
    sex_list  = [ 'female', 'male' ]
    cov_list  = [ 'haqi', 'income' ]
    omega     = 0.02
    #
    # covariate_table
    covariate_table = list()
    for node_name in node_list :
        for sex in sex_list :
            for age in age_list :
                for time in time_list :
                    row = {
                        'node_name' : node_name ,
                        'sex'       : sex ,
                        'age'       : age ,
                        'time'      : time ,
                        'omega'     : omega ,
                        'haqi'      : haqi(  node_name, sex, age, time) ,
                        'income'    : income(node_name, sex, age, time) ,
                    }
                    covariate_table.append(row)
    covariate_table = at_cascade.csv.covariate_both(covariate_table)
    random.shuffle(covariate_table)
    #
    # same cov
    cov_same = at_cascade.csv.covariate_same(covariate_table)
    #
    # sex_list
    sex_list = sex_list + [ 'both' ]
    #
    # check that omega and haqi are same for all nodes and all sexes
    for cov_name in { 'omega', 'haqi' } :
        triple_other = None
        for node_name in node_list :
            for sex in sex_list :
                triple = (node_name, sex, cov_name)
                if cov_same[triple] == triple :
                        assert triple_other == None
                        triple_other = triple
        for node_name in node_list :
            for sex in sex_list :
                assert cov_same[ (node_name, sex, cov_name) ] == triple_other
    #
    # income
    # same for all nodes, different for each sex
    triple_other = dict()
    for node_name in node_list :
        for sex in sex_list :
            triple = (node_name, sex, 'income')
            if cov_same[triple] == triple :
                    assert sex not in triple_other
                    triple_other[sex] = triple
    for node_name in node_list :
        for sex in sex_list :
            assert cov_same[ (node_name, sex, 'income') ] == triple_other[sex]
    print('cov_same_xam: OK')
#
if __name__ == '__main__' :
    main()