Understanding Periodicity

What is Periodicity?

The periodicity is the size of each bar of data. For example, end-of-day data has a periodicity of one day. Intraday data has a periodicity reflecting the number of minutes of information contained in each bar, as well as the frequency with which those bars are delivered.

The periodicity of a data series appears as an icon next to the data in the Portfolio View. It is also displayed on the Modify Data Dialog: Overview page.

The periodicity of a data series is set based on the actual data that is imported into that series and cannot be changed. For example, if you import end-of-day for a stock, you cannot change that data to intraday data, weekly data, or monthly data. Instead, you would need to import the desired periodicity separately.

Displaying Multiple Periodicities In Charts and Spreadsheets

The periodicity of a chart or spreadsheet is based on the context it is being displayed in. The context is determined by the data series you select to display.

:    Example: If you select to display a chart for a data series with 5-minute periodicity, the chart will have a 5-minute periodicity even if you add data with other periodicities to that chart.

Ä    Note: This means that "weekly" and "monthly" charts cannot be displayed from end-of-day data. You can display multiple years of end-of-day data, but there is no way to create weekly or monthly bars from that information. Similarly, there is no way to create daily bars from intraday information. In each of these cases, you would simply import the data in the periodicity you are interested in viewing.

Data with other periodicities will automatically be synchronized so that the most recently available bar at that time is displayed. In other words, the will line up to reflect when the information becomes available.

:    Example: Data with a 10-minute periodicity is added to a 1-minute periodicity chart. The 9:59 through 10:08 1-minute bars will synchronize with the 9:50 10-minute bar since that bar contains data for 9:50 through 9:59. At 10:10, the 10:09 1-minute bar and the 10:00 10-minute bar will both become available, so those will synchronize together.

Because of this synchronization, the times associated with data displayed in a chart or spreadsheet of a different periodicity will not match the times when it is displayed in its own periodicity or another periodicity.

:    Example: Data with a 10-minute periodicity is displayed in 1-minute periodicity chart. The data synchronized to the 10:00 1-minute bar is the 9:50 10-minute bar (since the 10:00 10-minute bar won’t be available until after 10:09.) However, when the 10-minute data is displayed in its own 10-minute chart, the data associated with 10:00 is the 10:00 bar.

Using Multiple Periodicities In Field Inputs

Fields use the same logic as charts and spreadsheets when combining information from multiple periodicities. In other words, data is synchronized so that the most recently known value is used with each bar.

:    Example: A field from data with a 10-minute periodicity is used as an input to a prediction for 1-minute data. The 9:59 through 10:08 1-minute bars will use the 9:50 10-minute bar since that bar contains data for 9:50 through 9:59. At 10:10, the 10:09 1-minute bar and the 10:00 10-minute bar will both become available, so the 10:09 1-minute bar will use the 10:09 10-minute bar.

Change and percent change preprocessing for predictions is based on the periodicity of the field being calculated. Since the value associated with slower periodicity data does not change every bar, you should not use these types of preprocessing for slower periodicity data.

:    Example: In the above example, the value of the input with 10-minute periodicity would not change for 10 bars. Therefore, the change and percent change of the values during these bars would be 0.