Class TailPassProfile

A posterior profile. Arrays are defensive, frozen copies.

Index

Constructors

Properties

k parameters scores thresholds

Accessors

covariance mean question_count question_mean questionCount questionMean std

Methods

at_k atK ci discovery linear linear_ci linearCi moment moment_ci momentCi sample stability

Constructors

constructor

new TailPassProfile(
    k: number,
    thresholds: readonly number[],
    parameters: readonly (readonly number[])[],
    scores: readonly number[],
): TailPassProfile
Parameters
- k: number
- thresholds: readonly number[]
- parameters: readonly (readonly number[])[]
- scores: readonly number[]
Returns TailPassProfile
- Defined in eval/tailpass.ts:160

Properties

`Readonly`k

k: number

`Readonly`parameters

parameters: readonly (readonly number[])[]

`Readonly`scores

scores: readonly number[]

`Readonly`thresholds

thresholds: readonly number[]

Accessors

covariance

get covariance(): number[][]
Returns number[][]
- Defined in eval/tailpass.ts:230

mean

get mean(): number[]
Returns number[]
- Defined in eval/tailpass.ts:224

question_count

get question_count(): number
Returns number
- Defined in eval/tailpass.ts:175

question_mean

get question_mean(): number[][]
Returns number[][]
- Defined in eval/tailpass.ts:221

questionCount

get questionCount(): number
Returns number
- Defined in eval/tailpass.ts:172

questionMean

get questionMean(): number[][]
Returns number[][]
- Defined in eval/tailpass.ts:212

std

get std(): number[]
Returns number[]
- Defined in eval/tailpass.ts:252

Methods

at_k

at_k(k: number, options?: Pick<TailPassOptions, "thresholds">): TailPassProfile
Parameters
- k: number
- options: Pick<TailPassOptions, "thresholds"> = {}
Returns TailPassProfile
- Defined in eval/tailpass.ts:345

atK

atK(k: number, options?: Pick<TailPassOptions, "thresholds">): TailPassProfile
Parameters
- k: number
- options: Pick<TailPassOptions, "thresholds"> = {}
Returns TailPassProfile
- Defined in eval/tailpass.ts:334

ci

ci(confidence?: number, options?: TailPassCIOptions): ProfileSummary
Parameters
- confidence: number = 0.95
- options: TailPassCIOptions = {}
Returns ProfileSummary
- Defined in eval/tailpass.ts:385

discovery

discovery(): [number, number]
Returns [number, number]
- Defined in eval/tailpass.ts:328

linear

linear(weights: readonly number[]): [number, number]
Exact mean/std, including cross-threshold covariance.
Parameters
- weights: readonly number[]
Returns [number, number]
- Defined in eval/tailpass.ts:278

linear_ci

linear_ci(
    weights: readonly number[],
    confidence?: number,
    options?: TailPassCIOptions,
): ScalarSummary
Parameters
- weights: readonly number[]
- confidence: number = 0.95
- options: TailPassCIOptions = {}
Returns ScalarSummary
- Defined in eval/tailpass.ts:443

linearCi

linearCi(
    weights: readonly number[],
    confidence?: number,
    options?: TailPassCIOptions,
): ScalarSummary
Parameters
- weights: readonly number[]
- confidence: number = 0.95
- options: TailPassCIOptions = {}
Returns ScalarSummary
- Defined in eval/tailpass.ts:403

moment

moment(lam: number): [number, number]
Integrate the full attainable score distribution, independently of the grid.
Parameters
- lam: number
Returns [number, number]
- Defined in eval/tailpass.ts:287

moment_ci

moment_ci(
    lam: number,
    confidence?: number,
    options?: TailPassCIOptions,
): ScalarSummary
Parameters
- lam: number
- confidence: number = 0.95
- options: TailPassCIOptions = {}
Returns ScalarSummary
- Defined in eval/tailpass.ts:450

momentCi

momentCi(
    lam: number,
    confidence?: number,
    options?: TailPassCIOptions,
): ScalarSummary
Parameters
- lam: number
- confidence: number = 0.95
- options: TailPassCIOptions = {}
Returns ScalarSummary
- Defined in eval/tailpass.ts:423

sample

sample(nDraws?: number, options?: SamplingOptions): TailPassDraws
Seeded draws are reproducible within JS, not bit-identical to NumPy.
Parameters
- nDraws: number = 4000
- options: SamplingOptions = {}
Returns TailPassDraws
- Defined in eval/tailpass.ts:352

stability

stability(): [number, number]
Returns [number, number]
- Defined in eval/tailpass.ts:331