import ast

import os

import json

import pandas as pd

import numpy as np

import random

import itertools as it

import math

from scipy.stats import scoreatpercentile as satp

from datetime import datetime,timedelta

from kariba_settings import hfca_2_pop, cc_correction_factor, get_fold_bins, cc_agg_fold, cc_agg_period, fold_start_date, fold_end_date, calib_node_pop, cluster_2_cc, cc_sim_start_date, cc_ref_start_date, cc_ref_end_date, tags_report_data_file, fit_terms_file, sim_data_dir, cc_penalty_model, fit_terms_types, hfca_id_2_cluster_ids, get_hfca_ids, get_cc_cluster_weight_factor, cluster_2_pops, sample_size_percentile, weighted_ccs_by_hfca_id_file, min_num_combos_per_hfca, load_prevalence_based_ccs, sample_size, cluster_2_mean_pop, get_cluster_category

from utils import warn_p, debug_p, feature_scale

from copy import deepcopy

def get_cc_penalty(fit_entry):

return None

def sim_key_2_group_key(sim_key):

return group_key

def get_sim_key(temp_h, const_h, itn_level, drug_level):

return str(temp_h) + '_' + str(const_h) + '_' + get_sim_group_key(itn_level, drug_level)

def get_sim_group_key(itn_level, drug_level):

return str(itn_level) + '_' + str(drug_level)

def sim_meta_2_itn_level(meta):

itn_level_struct = ast.literal_eval(meta['add_ITN_mult'])

itn_level = itn_level_struct[0][1][0][0][1]

return itn_level

def sim_meta_2_drug_cov(meta):

drug_coverage_level_struct = ast.literal_eval(meta['add_drug_multi_campaigns'])

drug_coverage_level = drug_coverage_level_struct[0][1][0]['coverage']

return drug_coverage_level

def sim_meta_2_temp_h(meta):

x_temp_h = float(meta['x_Temporary_Larval_Habitat'])

return x_temp_h

def sim_meta_2_const_h(meta):

const_h_struct = ast.literal_eval(meta['scale_larval_habitats_single'])

const_h = float(const_h_struct[0][1][1])

return const_h

def get_model_params(model):

return temp_h, const_h, itn_level, drug_coverage_level

def combine_tags_reports(base_dirs, output_dir):

print "DONE"

def cc_data_aggregate(model_clinical_cases, cluster_id, cc_cluster_weight_factor = None):

return ccs_model_agg, ccs_ref_agg, fold_bins

def update_fold_bins(date, cases_model, cases_ref, fold_bins):

return fold_bins

def fold_nan_clean(fold_bins):

return ccs_folded_model, ccs_folded_ref

def cc_data_nan_clean(ccs_model_agg, ccs_ref_agg, cluster_id):

return ccs_model_agg_clean, ccs_ref_agg_clean

def get_cc_model_ref_traces(model_clinical_cases, cluster_id, cc_cluster_weight_factor = None):

return ccs_model_agg, ccs_ref_agg

def load_fit_terms(fit_terms_file_path):

return fit_terms

# supply fit terms type that containing the fit term hierarchy path in fit_terms.json

# e.g. clinical cases penalty computed using rank correlation on folded clinical cases timeseries

# would have type ['cc_penalty', 'corr_folded', 'penalty']

def normalize_fit_terms(fit_terms_file_path):

return fit_terms

def unroll_term(fit_entry, path):

return unroll_term(fit_entry, path)

# can be implemented better w/ recursion

def roll_term(fit_term, path, extremal_term):

return fit_term

def gazetteer_list_update(data, file_path):

json.dump(cur_data, f_p, indent = 4)

def error_loading_fit_terms():

raise ValueError('Could not load fit terms! Check whether fit terms file at ' + os.path.join(sim_data_dir, fit_terms_file) + ' is accessible.')

def get_prevalence_opt_region_sims(best_fits, all_fits, cluster_id):

return opt_region_sims

def get_prevalence_based_cc(best_fits, all_fits, calib_data):

return weighted_ccs_model_agg_by_hfca

def weighted_ccs_combos(hfca_id, weighted_ccs_model_agg_by_hfca):