A time series upload field is a custom field that can capture and analyze the time series data output from devices such as GPS units. Other types of time series data can be from heart rate monitors, power meters, force plates, linear transducers and accelerometers. Using a time series upload field allows for the upload of this raw data exported from the device into the event form. Once uploaded, this data can be referenced by table duration calculations or table calculations in the form to work out information such as total session time, maximum and minimum values, time spent in different heart rate zones and other variables.
On normal forms when you enter data into a field, you’ll see the calculations fill out automatically. Calculations that reference a time series upload field aren’t automatic: you have to save the form and wait for time series processing. This data can take a while to process and the processing time can be slowed by server load. There is a form advanced property called Send emails after time series processing that will notify the user by email when the time series data has finished processing; however, most builders choose to turn this off.
Once the form has been saved and the time series data has finished processing, the user will generally use the performance history tool to view information about that form. For this reason, all of key fields need their advanced properties set to show in tables so you can see them in the performance history when they’ve finished calculating.
Once on the server, Smartabase will create a number of additional columns for the file. For every column in the file you uploaded, Smartabase will create a new column with the delta value for that variable. For example, let’s say you have a file with three columns: Time, Distance, Speed. Once uploaded you will have 6 columns: Time, Time.delta, Distance, Distance .delta, Speed, Speed.delta. Each delta column will calculate the current row minus the previous row. These delta columns may be referenced in calculations later in the form.
It is also important, when building a form that uses time series data, to ensure that the format of the upload file matches the format for that field and that the format of subsequent fields is the same as well. For example, if the raw data is formatted as mm:ss:S, then the time series upload field and subsequent table duration calculations or table calculations also need to use this format.
The type of time that the data is recorded against in the raw data can affect what type of calculation you should use. For example, total time vs interval time: in the first case, each record has a time value that records the amount of time elapsed since the start of the series but in the second case, each record is entered at regular time intervals. This affects how you would add up, for example, time spent in heart rate zones. If you are working with interval time, you wouldn’t be able to use a table duration calculation. Instead, you would use the countifpositive aggregate type in a normal table calculation, then a normal duration calculation (using same format as previous calculation) that returns the result of the previous calculation multiplied by the sampling rate. To reference any column in the uploaded time series file, you will need to use a specific naming structure. Use the name that you gave the time series upload field with the column heading from the file that you want to reference. For example, if the time series upload field is named GPS and you are referencing a column named Distance, you would write it as GPS.distance in the calculation.
Steps to build a form that uses a time series upload field
- Log into the builder interface.
- Select the Event forms tool.
- Select New event form from the Options drop-down.
- Name and categorize the event form.
- Set form advanced properties, including setting Send emails after time series processing to False, if needed.
- Add a Time series upload field from the custom menu.
- Name the time series upload field. Avoid using a name that already exists as a variable in the raw data spreadsheet.
- Set field advanced properties, including:
- Format to match input data format.
- Skip rows if you want to ignore a specific number of rows at the start of the CSV from any subsequent calculations. This will apply to all files for all athletes so it is only recommended if you are using devices which have a known period of warming up, or trial runs.
- Take every Nth row will take every x rows as actual data (e.g. every 4th or 5th row). Use this to ignore some data at high recording frequencies. This will apply to all files for all athletes.
- Data format to select the format of the CSV file. Smartabase is compatible with the format of the export files from Catapult GPS devices. If you are using another technology then leave blank or select Generic.
- Header override to optionally replace the header in the time series file. Headers must be comma separated and contain no special characters. If you would like to replace the names of the column headings or if the file you are adding doesn’t have a header row, use this property to define your desired headers.
Steps to calculate session time from a time series upload field
- Add a Table duration calculation field from the calculations menu.
- Set field advanced properties as follows:
- Format: input data format.
- Auto calculate: Yes.
- Default show in tables: Yes.
- Write a table duration calculation as follows:
- Aggregate: set to Max if positive.
- Reference the time series upload field as field name.column name.
- Add .Delta (because output is a time).
- Save the event form.