NavAbilitySDK

mutable struct Categorical <: Distribution

Categorical distribution specified by a set of probabilities summing up to 1.

struct Client

The context for a session, made from a user, robot, and session. Users can have multiple robots and robots can have multiple sessions. So this indicates a unique session.

DevNotes

• Rename to SessionKey

abstract type Distribution

Abstract parent type for all distributions.

mutable struct FullNormal <: Distribution

Multidimensional normal distribution specified by means and a covariance matrix.

struct GraphVizApp

Helper type for linking to App visualization of a factor graph for user:robot:session.

abstract type InferenceType

Abstract parent type for all InferenceTypes, which are the functions inside of factors.

struct MapVizApp

Helper type for linking to App visualization of geometric map for user:robot:session.

struct MixtureInferenceType <: NavAbilitySDK.InferenceType

InferenceType for MixtureData.

mutable struct Normal <: Distribution

One dimensional normal distribution.

struct Pose2AprilTag4Corners <: NavAbilitySDK.InferenceType

InferenceType for Pose2AprilTag4CornersData.

struct Pose2Point2BearingRange <: NavAbilitySDK.InferenceType

Pose2Point2BearingRangeInferenceType is used to represent a bearing

• range measurement.

mutable struct Rayleigh <: Distribution

One dimensional Rayleigh distribution.

struct Scope

Some calls interact across multiple users, robots, and sessions. A scope allows you to specify these more complex contexts.

struct SolveOptions

Solver options including the solve key and whether the parametric solver should be used.

mutable struct Uniform <: Distribution

One dimensional uniform distribution.

struct ZInferenceType <: NavAbilitySDK.InferenceType

ZInferenceType is used by many factors as a common inference type that uses a single distribution to express a constraint between variables. Used by: Prior, LinearRelative, PriorPose2, PriorPoint2, Pose2Pose2, Point2Point2Range, etc.

LinearRelativeData(; Z, kwargs...)

Create a ContinousScalar->ContinousScalar (also known as Pose1->Pose1) factor with a distribution Z representing the 1D relationship between the variables, e.g. Normal(1.0, 0.1).

Default value of Z = Normal(1.0, 0.1).

MixtureData(mechanics, components, probabilities, dims)

Create a Mixture factor type with an underlying factor type, a named set of distributions that should be mixed, the probabilities of each distribution (the mix), and the dimensions of the underlying factor (e.g. ContinuousScalar=1, Pose2Pose2=3, etc.).

Args: mechanics (Type{FactorData}): The underlying factor data type, e.g. Pose2Pose2Data. NOTE: This will change in later versions but for now it can be any of the FactorData classes (e,g, LinearRelative, not the object LinearRelative()). components (NamedTuple): The named tuple set of distributions that should be mixed, e.g. NamedTuple(hypo1=Normal(0, 2)), hypo2=Uniform(30, 55)). probabilities (List[float]): The probabilities of each distribution (the mix), e.g. [0.4, 0.6]. dims (int): The dimensions of the underlying factor, e.g. for Pose2Pose2 it's 3.

NavAbilityHttpsClient()
NavAbilityHttpsClient(apiUrl; auth_token, authorize)

Main interactions with API for queries and mutations go through here.

DevNotes

• TODO TBD, rather use upstream retry logic instead,
  • see https://github.com/JuliaWeb/HTTP.jl/pull/974/files

Point2Point2RangeData(; range, kwargs...)

Create a Point2->Point2 range factor with a 1D distribution:

• range: The range from the pose to the point, default Normal(1, 1).

Pose2AprilTag4CornersData(
    id,
    corners,
    homography;
    K,
    taglength,
    kwargs...
)

Create a AprilTags factor that directly relates a Pose2 to the information from an AprilTag reading. Corners need to be provided, homography and tag length are defaulted and can be overwritten.

Pose2Point2BearingRangeData(; bearing, range, kwargs...)

Create a Pose2->Point2 bearing+range factor with 1D distributions:

• bearing: The bearing from the pose to the point, default Normal(0, 1).
• range: The range from the pose to the point, default Normal(1, 1).

Pose2Pose2Data(; Z, kwargs...)

Create a Pose2->Pose2 factor with a distribution Z representing the (x,y,theta) relationship between the variables, e.g. FullNormal([1,0,0.3333*π], diagm([0.01,0.01,0.01])).

Default value of Z = FullNormal([1,0,0.3333*π], diagm([0.01,0.01,0.01])).

Pose3Pose3Data(; Z, kwargs...)

Create a Pose3->Pose3 factor with a distribution Z representing the (x,y,theta) relationship between the variables, e.g. FullNormal([1;zeros(5)], diagm(0.01*ones(6))).

Default value of Z = FullNormal(zeros(6), diagm(0.01*ones(6))).

Pose3Pose3RotationData(; Z, kwargs...)

Create a partial factor on Rotation only on Pose3->Pose3 with a distribution Z representing the relationship between the variables.

Default value of Z = FullNormal(zeros(3), diagm(0.01*ones(3))).

PriorData(; Z, kwargs...)

Create a prior factor for a ContinuousScalar (a.k.a. Pose1) with a distribution Z representing 1D prior information, e.g. Normal(0.0, 0.1).

Default value of Z = Normal(0.0, 0.1).

PriorPoint2Data(; Z, kwargs...)

Create a prior factor for a Point2 with a distribution Z representing (x,y) prior information, e.g. FullNormal([0.0, 0.0.0], diagm([0.01, 0.01])).

Default value of Z = FullNormal([0.0, 0.0], diagm([0.01, 0.01])).

PriorPose2Data(; Z, kwargs...)

Create a prior factor for a Pose2 with a distribution Z representing (x,y,theta) prior information, e.g. FullNormal([0.0, 0.0, 0.0], diagm([0.01, 0.01, 0.01])).

Default value of Z = FullNormal([0.0, 0.0, 0.0], diagm([0.01, 0.01, 0.01])).

PriorPose3Data(; Z, kwargs...)

Create a prior factor for a Pose3 with a distribution Z representing (x,y,z,i,j,k) prior information, e.g. FullNormal(zeros(6), diagm(0.01*ones(6))).

Default value of Z = FullNormal(zeros(6), diagm(0.01*ones(6))).

ScatterAlignPose2Data(varType, cloud1, cloud2)
ScatterAlignPose2Data(varType, cloud1, cloud2, bw1)
ScatterAlignPose2Data(
    varType,
    cloud1,
    cloud2,
    bw1,
    bw2;
    mkd1,
    mkd2,
    kw_sap,
    kwargs...
)

Returns <:FactorData

Internal utility function to create the correct solver data (variable data) given a variable type.

addVariable Add a variable to the NavAbility Platform service Example

addVariable(client, context, "x0", NVA.Pose2)

createDownloadEvent(navAbilityClient, userId, fileId)

Request URLs for data blob download.

Args: navAbilityClient (NavAbilityClient): The NavAbility client. userId (String): The userId with access to the data. fileId (String): The unique file identifier of the data blob.

createUploadEvent(navAbilityClient, filename, filesize)
createUploadEvent(
    navAbilityClient,
    filename,
    filesize,
    parts
)

Request URLs for data blob upload.

Args: navAbilityClient (NavAbilityClient): The NavAbility client. filename (String): file/blob name. filesize (Int): total number of bytes to upload. parts (Int): Split upload into multiple blob parts, FIXME currently only supports parts=1.

getStatusLatestEvent(navAbilityClient, id)

Get the latest status message for a request.

Args: navAbilityClient (NavAbilityClient): The NavAbility client. id (String): The ID of the request that you want the latest status on.

getStatusMessagesEvent(navAbilityClient, id)

Get all the statuses for a request.

Args: navAbilityClient (NavAbilityClient): The NavAbility client. id (String): The ID of the request that you want the statuses on.

getStatusesLatest(navAbilityClient, ids)

Helper function to get a dictionary of all latest statues for a list of results.

Args: navAbilityClient (NavAbilityClient): The NavAbility client. ids (Vector{String}): A list of the IDS that you want statuses on.

incrDataLabelSuffix

If the blob label thisisme already exists, then this function will return the name thisisme_1. If the blob label thisisme_1 already exists, then this function will return the name thisisme_2.

DO NOT EXPORT, Duplicate functionality from DistributedFactorGraphs.jl.

natural_lt(x, y)

Natural less than for sorting,

julia sort(["x10"; "x1", "x11"]; lt=NavAbilitySDK.natural_lt)`

Notes

• duplicated from DFG, hence don't export

waitForCompletion(
    navAbilityClient,
    requestIds;
    maxSeconds,
    expectedStatuses,
    exceptionMessage
)

Wait for the requests to complete, poll until done.

Args: requestIds (List[str]): The request IDs that should be polled. maxSeconds (int, optional): Maximum wait time. Defaults to 60. expectedStatus (str, optional): Expected status message per request. Defaults to "Complete".