When configuring data ingestion from a feed or data source, ArcGIS Analytics for IoT allows you to specify the date and time properties of the data.
Specify date and time configuration
The date and time properties of a feed or data source can be specified on the Identify Key Fields step under the Date & Time section.
Date & Time
The features in a feed or data source may or may not have time fields available.
It is not required to set a Start Time, End Time, or Date Format to analyze and process data. However, some tools in real-time and big data analytics require a start time or a start time and end time to be defined in order to perform temporal analysis.
Additionally, if date information is represented as strings in your input data, a date format is required for Analytics for IoT to successfully parse this string into a date field.
There are three time configurations available:
- Neither Start Time or End Time fields are specified. The data has no representation of time, for example, a country boundaries dataset without any date or time fields. Simply leave these parameters empty.
- A Start Time field is specified. The data records represent their state as of that specific point in time, also known as instant time. For example, sensor observations which record the time of each reading.
- Both Start Time and End Time fields are specified. The data records represent the state of that feature or observation for that duration of time, also known as interval time. For example, thunderstorm forecast warning polygons with a start and end time of each warning.
Date format - epoch time
If your data represents dates with epoch time values (also known as UNIX time), you do not need to specify a date format. Epoch time is a system for describing a point in time as the number of seconds since 00:00:00 Thursday, 1 January 1970 (UTC). This can be specified as seconds since or milliseconds since this point in time. Example values of epoch time that could be processed by Analytics for IoT without specifying a date format are 1563997859 or 1563997859000.
Date format - strings
Dates and datetimes stored in a string field can be represented in many different ways. For example, the following strings refer to the same point in time, but they are represented differently:
- Aug 7, 2019 7:02:01
If you have fields that express date values as strings, you will need to specify a date format so that Analytics for IoT knows how to parse this string into a date. This is done by providing a datetime formatting string in the Date Format parameter. Below are several examples of formatting datetimes:
- 12/27/2015 16:39:20
- MM/dd/yyyy HH:mm:ss
Analytics for IoT supports time up to milliseconds. Any date strings with further precision will need to defined as in the example above.
The complete list of elements that can be used to specify the date format for strings can be found in the table below. Additional information regarding the configuration of a date time formatting string is available for the corresponding Joda time format DateTimeFormat class that is used to convert strings representing dates into date objects.
|Date format string symbol||Meaning||Examples|
century of era (>=0)
year of era (>= 0)
week of weekyear
day of week
day of week
day of year
month of year
July; Jul; 07
day of month
halfday of day
hour of halfday (0~11)
clockhour of halfday (1~12)
hour of day (0~23)
clockhour of day (1~24)
minute of hour
second of minute
fraction of second
Pacific Standard Time; PST
time zone offset/id
-0800; -08:00; America/Los_Angeles
Date format - Esri date fields
If your data has date or date and time information stored in a date field, you do not need to specify a date format. Feature layers, shapefiles, and Parquet files can hold date fields that store date or date and time values. Analytics for IoT can load date and time values from such fields without any additional properties or settings needed.