Stream visualization server

Streamvis project is a webserver and a collection of apps for visualization of data streams. It is based on bokeh and generally works with zmq streams.

An example of user application:

Build

The build is triggered upon pushing a tag into a master branch and involves running the Github Actions script, which builds a package and uploads it to paulscherrerinstitute anaconda channel. A tagged release commit can be created with make_release.py script.

To build a local conda package without uploading it to the anaconda channel:

$ conda build ./conda-recipe

A docker image can be built from the latest version on paulscherrerinstitute anaconda channel:

$ docker build ./docker

Install

The default installation procedure:

$ conda install -c paulscherrerinstitute streamvis

If a package was built locally:

$ conda install --use-local streamvis

Running the server

$ streamvis <app_name> [--opt-params]

Navigate to http://localhost:5006/ in your browser to start using the streamvis application.

To get a general help, a list of available applications and optional parameters:

$ streamvis -h

Message metadata entries

Required metadata entries:

• type: str - image data type, e.g. "uint16", or "float32"
• shape: Iterable[int] - image shape in pixels

Optional jungfrau-related metadata entries:

• detector_name: str - detector name (e.g. "JF01T03V01"), required for adc->keV conversion and geometry
• gain_file: str - path to a gain file, required for adc->keV conversion
• pedestal_file: str - path to a pedestal file, required for adc->keV conversion and mask
• daq_rec: int - a last bit is used to determine whether detector is in a highgain mode
• module_map: Iterable[int] - a mapping between data regions and detector module positions (e.g. [0, -1, 1] - a second module is switched off)
• mask: bool - superseed a user GUI selection for mask
• gap_pixels: bool - superseed a user GUI selection for gap_pixels
• geometry: bool - superseed a user GUI selection for geometry
• double_pixels: str - superseed a user GUI selection for double_pixels, can be "keep", "mask", or "interp"

Statistics tab:

• pulse_id: int - is required for statistics to be collected, data is grouped into runs based on this value
• is_good_frame: bool - causes a metadata issue if not True, increments a number in "Bad Frames" column
• saturated_pixels: int - causes a metadata issue if not 0, increments a number in "Sat pix frames" column
• laser_on: bool - is used to split statistics between laser_on/laser_off columns, those columns are hidden if this key is not present
• is_hit_frame: bool - increments a number in "Laser ON/OFF frames" columns (those columns are hidden if laser_on is not present)

Hitrate tab:

• pulse_id: int - is required for statistics to be collected, determines which point along x-axis is updated
• is_hit_frame: bool - determines whether it's a hit (also whether the data may be displayed when "Show Only Hits" button is toggled)

Radial Profile tab:

• pulse_id: int - is required for statistics to be collected, the corresponding data is ignored if values is not within "Pulse ID Window" from the most recent pulse_id received
• radint_q: Iterable[float] - a vector of q values (x-axis)
• radint_I: Iterable[float] - a vector of intensity values (y-axis)
• laser_on: bool - determines which graph from "Frames laser on/off" is being updated

ROI intensities tab / Intensity ROIs overlay:

• roi_x1: Iterable[float] - a vector of intensity ROI left borders
• roi_x2: Iterable[float] - a vector of intensity ROI right borders
• roi_y1: Iterable[float] - a vector of intensity ROI bottom borders
• roi_y2: Iterable[float] - a vector of intensity ROI top borders
• roi_intensities_normalised: Iterable[float] - a vector of corresponding ROI intensities

Saturated Pixels overlay:

• saturated_pixels_x: Iterable[float] - x-coordinates of saturated pixels (attempt to derive if not present in case of raw data (uint16))
• saturated_pixels_y: Iterable[float] - y-coordinates of saturated pixels (attempt to derive if not present in case of raw data (uint16))
• saturated_pixels: int - a number of saturated pixels (attempt to derive if not present in case of raw data (uint16))

Spots overlay:

• spot_x: Iterable[float] - x-coordinates of spots
• spot_y: Iterable[float] - y-coordinates of spots
• number_of_spots: int - should be equal to a length of both, spot_x and spot_y

Resolution Rings overlay:

• detector_distance: float - distance to detector in meters
• beam_energy: float - beam energy in eV
• beam_center_x: float - beam x-coordinate
• beam_center_y: float - beam y-coordinate

Disabled Modules overlay (requires a valid detector_name):

• disabled_modules: Iterable[int] - indexes of modules to display as disabled

Aggregated images:

• aggregated_images: int - in case of aggregation, treat a received image as a sum of that number of images