Time periods in eCalc™
eCalc™ does calculations between a global start date and a global end date. This global time period will be split into many smaller time periods, where each of them follow immediately after each other, based on other dates encountered in the YAML file and in the various time series resource files.
Finding the global start and end dates
The global start and end dates can be specified directly in the YAML file. eCalc™ will first check if the global start and end dates are specified by the users. This is done using the START and END keywords in the YAML file. If the START and END keywords are not specified, the dates will be determined by the time series data given in TIME_SERIES. The earliest date and latest date encountered within the time series data with INFLUENCE_TIME_VECTOR set to True will be considered the global start date and global end date, respectively.
Any dates defined in the YAML file (e.g. in ENERGY_USAGE_MODEL) will not influence the global start and end dates.
Splitting the global time period into smaller time periods
The period of time between the global start and end dates, will be split into many smaller time periods. This is done by finding:
- all other dates in all the time series data that are between the global start and end dates, and has INFLUENCE_TIME_VECTOR set to True.
- all dates defining temporal models in the YAML file (e.g. in ENERGY_USAGE_MODEL, FUEL or VARIABLES) that are between the global start and end dates.
When all N dates are found and sorted, (N-1) time periods will be generated. The first time period will start at the global start date and end at the next date in the sorted list. The second time period will start where the previous ends (start is inclusive, end is exclusive), and so on until the last time period which will end at the global end date.
The time series input given in the TIME_SERIES may not have values for all the generated periods. Any missing values will be found using the chosen INTERPOLATION_TYPE.
Be aware that when a time series has INFLUENCE_TIME_VECTOR set to False, any dates found in the csv file that are not already included in the global time vector, will be left out.
The last row given for the variables in a single csv file is considered to be the values for the variables in the period starting at that date. If no later dates are given in any other csv files, or no END date is given, the last row in that file is in reality irrelevant - it will not be used for anything. But, if an end date is defined through the END keyword, that last row will be the values for the variables in that additional added period of time.
If one csv file ends several periods of time before another csv file, the last row in the first csv file will be extrapolated. Either just for one period (if EXTRAPOLATION is False), or for all remaining periods (if EXTRAPOLATION is True).