Dis-aggregation
Basic and Dis-aggregated Series
The original series, i.e. the series to be dis-aggregated, is called basic series. The series with the dis-aggregates is called the dis-aggregated series.
In dis-aggregating series following options exist:
1. for instantaneous observations
- basic and dis-aggregated series data are equal,
- dis-aggregated series data are interpolated linearly between the mid-point values of the basic series data;
2. for accumulative observations
- dis-aggregated series data are fractions of the basic series data; if there are n dis-aggregated series elements in a basic series interval, then the fraction is 1/n: dis-aggregated series data = 1/n * basic series data,
- as above, but additionally with linear interpolation between the midpoints of the basic series intervals.
Series Codes
The series to dis-aggregate can be selected or de-selected by clicking the checkbox of the series. By clicking your right mouse button on top of the Series list box you can select or de-select all series.
Interpolation
When the 'Interpolation' checkbox is checked, the dis-aggregated series data are interpolated linearly.
To execute dis-aggregation following entries/selections have to be made:
- Select one or more basic series from the Series Codes list.
- Select time intervals of aggregated series where Dt~basic~ > Dt~dis-aggregated~
- in addition you have to indicate whether the dis-aggregates should be interpolated linearly between the mid-points of the basic series or should be taken as constant over a basic series interval.
- Press <Execute> to execute the dis-aggregation.
The newly aggregated series can be shown in a Graph, Viewed within the spreadsheet viewer, Saved to the database or Exported.
Limitations
The step between the time interval units for aggregation and dis-aggregation should generally not be more than one; e.g. aggregation from hour to day is allowed but aggregation from hour to month is not allowed. An exception applies for the cases where the basic time unit of the resulting series equals the time interval unit of the basic series, e.g. from day to week. Aggregations and dis-aggregations which are not allowed are blocked by the menu.
Influence of the processing period while aggregating
While aggregating time series the start period is very important. Some examples will be given:
Aggregating hourly values to daily values. If the processing period is set from 1-1-1986 00:00:00 to 1-1-1987 00:00:00, the hourly values from 00:00:00 to 23:00:00 will be aggregated to daily values. If the processing period is set from 1-1-1986 08:00:00 to 1-1-1987 08:00:00, the hourly values from 08:00:00 to 07:00:00 of the next day will be aggregated to daily values.
Aggregating daily values to monthly values. If the processing period is set from 1-1-1986 to 1-1-1987, the daily values from the first day of the month till the last day of the month will be aggregated to monthly values. If the processing period is set from 12-1-1986 to 12-1-1987, the daily values from day 12 till day 11 of the next month will be aggregated to monthly values.
