Skip to main content
Version: 7.9

system.tag.queryTagHistory

This function is used in Python Scripting.

Description

Issues a query to the Tag Historian. Querying tag history involves specifying the tags and the date range, as well as a few optional parameters. The Tag historian will find the relevant history and then interpolate and aggregate it together into a coherent, tabular result set.

This is useful when you're trying to retrieve tag history data over a period of time (i.e., multiple timeslices over a period of time). If you are trying to take a range of time and aggregate the data to a single value then consider using system.tag.queryTagCalculations.

This function takes a list of strings, where each string is a tag path, like Tanks/Tank5 or `[OracleProvider]Sump/Out2``. See also: Tag Paths.

The return size determines how the underlying data is aggregated and/or interpolated. If a distinct return size is specified, that will be the number of rows in the resulting dataset. The special numbers 0 and -1 mean "Natural" and "On-Change", respectively. "Natural" calculates a return size based on the rate of the logging historical scan classes. For example, if you query 1 hour of data for a scan class logging every minute, the natural return size is 60. "On-Change means that you'll get an entry whenever any of the tags under consideration have changed.

Instead of defining a fixed return size, the parameters intervalHours and intervalMinutes can be used. These parameters can be used independently or together to define a "window size". For example, if you defined a 1 hour range, with intervalMinutes=15, you would get 4 rows as a result.

The span of the query can be specified using startDate and endDate. You can also use rangeHours and rangeMinutes in conjunction with either start or end date to specify the range in dynamic terms. For example, you could specify only "rangeHours=-8" to get the last 8 hours from the current time. Or you could use "startDate='2012-05-30 00:00:00', rangeHours=12" to get the first half of the day for May 30th, 2012. The aggregation mode is used when the data is denser than what you asked for. This happens when using fixed return sizes, as there will often be multiple raw values for the window interval defined. Another common operation is to set the return size to 1, in order to use these aggregate functions for calculation purposes. The available functions are:

  • "MinMax" - will return two entries per time slice - the min and the max.
  • "Average" - will return the time-weighted average value of all samples in that time slice.
  • "LastValue" - returns the most recent actual value to the end of the window. Note that if a value does not exist in this window, a 0 will be returned in cases where - interpolation is turned off.
  • "SimpleAverage" - returns the simple mathematical average of the values - ((V1+V2+...+Vn)/n)
  • "Maximum" - the maximum value of the window.
  • "Minimum" - the minimum value of the window.
  • "DurationOn" - the time, in seconds, that a value has been boolean true
  • "DurationOff" - the time, in seconds, that a value has been boolean false
  • "CountOn" - the number of times the value has transitioned to boolean true
  • "CountOff" - the number of times the value has transitioned to boolean false
  • "Count" - the number of "good", non-interpolated values per window.
  • "Range" - the difference between the min and max
  • "Variance" - the variance for "good", non-interpolated values. Does not time weight.
  • "StdDev" - the standard deviation for "good", non-interpolated values. Does not time weight.
  • "PctGood" - the percentage of time the value was good.
  • "PctBad" - the percentage of time the value was bad.

Client Permission Restrictions

This scripting function has no Client Permission restrictions.

Syntax

tip

This function accepts keyword arguments.

system.tag.queryTagHistory(paths, startDate, endDate, returnSize, aggregationMode, returnFormat, columnNames, intervalHours, intervalMinutes, rangeHours, rangeMinutes, aggregationModes, includeBoundingValues, validateSCExec, noInterpolation, ignoreBadQuality)

Parameters

TypeParameterDescription
PySequencepathsAn array of tag paths (strings) to query. Each tag path specified will be a column in the result dataset.
DatestartDateThe earliest value to retrieve. If omitted, 8 hours before current time is used.
DateendDateThe latest value to retrieve. If omitted, current time is used.
IntegerreturnSizeThe number of samples to return. -1 will return values as they changed, and 0 will return the "natural" number of values based on the logging rates of the scan class(es) involved. -1 is the default.
StringaggregationModeThe mode to use when aggregating multiple samples into one time slice. Valid values are: "Average" (time-weighted), "MinMax", "LastValue", "SimpleAverage", "Sum", "Minimum", "Maximum", "DurationOn", "DurationOff", "CountOn", "CountOff", "Count", "Range", "Variance", "StdDev", "PctGood", and "PctBad". Default is "Average" (time-weighted).
StringreturnFormatUse "Wide" to have a column per tag queried, or "Tall" to have a fixed-column format. Default is "Wide".
PySequencecolumnNamesAliases that will be used to override the column names in the result dataset. Must be 1-to-1 with the tag paths. If not specified, the tag paths themselves will be used as column titles.
IntegerintervalHoursAllows you to specify the window interval in terms of hours, as opposed to using a specific return size.
IntegerintervalMinutesSame as intervalHours, but in minutes. Can be used on its own, or in conjunction with intervalHours.
IntegerrangeHoursAllows you to specify the query range in hours, instead of using start and end date. Can be positive or negative, and can be used in conjunction with startDate or endDate.
IntegerrangeMinutesSame as rangeHours, but in minutes.
PySequenceaggregationModesA one-to-one list with paths specifying an aggregation mode per column.
BooleanincludeBoundingValuesA boolean flag indicating that the system should attempt to include values for the query bound times if possible. See Seeded Values for more information.
BooleanvalidateSCExecA boolean flag indicating whether or not data should be validated against the scan class execution records. If false, data will appear flat (but good quality) for periods of time in which the system wasn't running. If true, the same data would be bad quality during downtime periods.
BooleannoInterpolationA boolean flag indicating that the system should not attempt to interpolate values in situations where it normally would. This will also prevent the return of rows that are purely interpolated.
BooleanignoreBadQualityA boolean flag indicating that bad quality values should not be used in the query process. If set, any value with a "bad" quality will be completely ignored in calculations and in the result set.
IntegertimeoutTimeout in milliseconds for Client Scope. This property is ignored in the Gateway Scope. [optional]

Returns

Dataset - A dataset representing the historian values for the specified tag paths. The first column will be the timestamp, and each column after that represents a tag.

Scope

All

Code Examples

Example #1
# The following example will return a dataset with one row detailing maximum value of a tag named 'Sine' for the past 30 minutes.
endTime = system.date.now()
startTime = system.date.addMinutes(endTime, -30)
dataSet = system.tag.queryTagHistory(paths=['Sine'], startDate=startTime, endDate=endTime, returnSize=1, aggregationMode="Maximum", returnFormat='Wide')