Faery supports streams of various data types with different characteristics. Understanding these streams is crucial for understanding the behavior of Faery---and in particular the Python API.
7.1Stream types¶
All streams are implemented as a Stream which iterates over some data.
The data type characterizes what is contained in the stream.
We mostly operate with two types: events and frames:
| Stream | Data Type | Characteristics |
|---|---|---|
| EventsStream | np.ndarray | Sparse event data represented as timestamps, 2-d coordinates, and polarity bit (t, x, y, p). |
| FrameStream | Frame | Dense frame data represented as [timestamp, np.ndarray](python |
7.2Finite, infinite, and regular streams¶
Apart from the type of data, streams can have different characteristics that matter a great deal for what you can do with them. For instance, a finite stream can be processed in a single pass, while an infinite stream requires some kind of continuous processing. Additionally, streams can be structured in time by sending data at regular intervals.
| Stream type | Description | Examples |
|---|---|---|
| InfiniteStream | An infinite stream requires some kind of continuous processing. | Reading from a camera or UDP source. |
| FiniteStream | A finite stream can be processed in a single pass. | Reading from a file or a finite source. |
| RegularStream | A regular stream sends data at regular intervals. | Filtering an event stream to output events at regular intervals. |
| FiniteRegularStream | A finite regular stream sends data at regular intervals and can be processed in a single pass. | Filtering a finite event stream to output events at regular intervals. |