Creates a dataset in the right format (with values for all 16 states of the DELPHI model) for the Optimal Vaccine Allocation team

Generates the prediction datasets from the date with 100 cases and from the day of running, including columns containing Confidence I

Creates the final MECE policies in the US from the intermediary policies dataframe :param df_policies_US: intermediary dataframe with process

Creates the final MECE policies in the US from the intermediary policies dataframe :param df_policies_US: intermediary dataframe with process

Creates the final MECE policies in the US from the intermediary policies dataframe :param df_policies_US: intermediary dataframe with process

Creates the balancing coefficient (regularization coefficient between cases & deaths in cost function) as well as the cases and deaths data o

Creates the balancing coefficient (regularization coefficient between cases & deaths in cost function) as well as the cases and deaths data o

Creates the balancing coefficient (regularization coefficient between cases & deaths in cost function) as well as the cases and deaths data o

Processes the IHME policy data in the US to create the right intermediary features with the right names :param df_policy_raw_us: raw datafram

Processes the IHME policy data in the US to create the right intermediary features with the right names :param df_policy_raw_us: raw datafram

Processes the IHME policy data in the US to create the right intermediary features with the right names :param df_policy_raw_us: raw datafram

Generates the nested dictionary with all the policy predictions which will then be saved as a JSON file to be used on the website

Generates the nested dictionary with all the policy predictions which will then be saved as a JSON file to be used on the website

Generates the nested dictionary with all the policy predictions which will then be saved as a JSON file to be used on the website

Generates the nested dictionary with all the policy predictions which will then be saved as a JSON file to be used on the website

Generates the format of the dictionary that will compose the dataframe with all backtest metrics based on the get_mae and get_mse fla

Creates aggregations at the continent level as well as associated confidence intervals :param df_predictions: dataframe containing th

Creates aggregations at the continent level as well as associated confidence intervals :param df_predictions: dataframe containing th

Creates aggregations at the continent level as well as associated confidence intervals :param df_predictions: dataframe containing th

Creates aggregations at the continent level as well as associated confidence intervals :param df_predictions: dataframe containing th

Creates aggregations at the country level as well as associated confidence intervals :param df_predictions: dataframe containing the

Creates aggregations at the country level as well as associated confidence intervals :param df_predictions: dataframe containing the

Creates aggregations at the country level as well as associated confidence intervals :param df_predictions: dataframe containing the

Creates aggregations at the country level as well as associated confidence intervals :param df_predictions: dataframe containing the

Creates aggregations at the world level as well as associated confidence intervals :param df_predictions: dataframe containing the ra

Creates aggregations at the world level as well as associated confidence intervals :param df_predictions: dataframe containing the ra

Creates aggregations at the world level as well as associated confidence intervals :param df_predictions: dataframe containing the ra

Creates aggregations at the world level as well as associated confidence intervals :param df_predictions: dataframe containing the ra

Updates the backtest metrics dictionary with metrics values for that particular area tuple :param df_backtest: pre-processed datafram

Generates the lower and upper bounds of the past parameters used as warm starts for the optimization process to predict with DELPHI: the outp

Generates the lower and upper bounds of the past parameters used as warm starts for the optimization process to predict with DELPHI: the outp

Checks that there is enough historical and prediction data to perform the backtest based on the user input :param df_historical: a da

Generates a concatenation of all historical data available in the danger_map folder, all areas starting from the prediction date give

Computes MAPE on cases & deaths (averaged) either on last 15 days of historical data (if there are more than 15) or exactly the number of day

Computes MAPE on cases & deaths (averaged) either on last 15 days of historical data (if there are more than 15) or exactly the number of day

Computes MAPE on cases & deaths (averaged) either on last 15 days of historical data (if there are more than 15) or exactly the number of day

Computes the normalized policy shifts and the current policy in each area of the world except the US (done in a separate function) :param

Computes the normalized policy shifts and the current policy in each area of the world except the US (done in a separate function) :param

Computes the normalized policy shifts and the current policy in each state of the US :param policy_data_us_only: processed dataframe with the

Computes the normalized policy shifts and the current policy in each state of the US :param policy_data_us_only: processed dataframe with the

Retrieve the predicted data on the prediction_date given as an input by the user running :param prediction_date: prediction date to b

Returns actual cases data for the given country and province :param country: str, the name of the country :param province: s

Returns actual cases data for the given country and province :param country: str, the name of the country :param province: s

Returns actual cases data for the given country and province :param country: str, the name of the country :param province: s

Obtain the value of the loss function depending on the optimizer (as it is different for global optimization using simulated annealing) :

Obtain the value of the loss function depending on the optimizer (as it is different for global optimization using simulated annealing) :

Obtain the value of the loss function depending on the optimizer (as it is different for global optimization using simulated annealing) :

Obtain the value of the loss function depending on the optimizer (as it is different for global optimization using simulated annealing) :

:return: a DataFrame where the column of interest is 'testing_cnt_daily' which gives the numbers of new daily tests per state

Compute the Maximum Absolute Percentage Error between two lists :param y_true: list of true historical values :param y_pred:

Compute the Maximum Absolute Percentage Error between two lists :param y_true: list of true historical values :param y_pred:

Compute the Maximum Absolute Percentage Error between two lists :param y_true: list of true historical values :param y_pred:

Reads the policy data from the Oxford dataset online and processes it to obtain the MECE policies for all other countries than the US :pa

Reads the policy data from the Oxford dataset online and processes it to obtain the MECE policies for all other countries than the US :pa

Reads and processes the policy data from IHME to obtain the MECE policies defined for DELPHI Policy Predictions :param filepath_data_sandbox:

Reads and processes the policy data from IHME to obtain the MECE policies defined for DELPHI Policy Predictions :param filepath_data_sandbox:

Saves the parameters and predictions datasets (since 100 cases and since the day of running) based on the different flags and the inp