al6x

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
from matplotlib.ticker import FuncFormatter, LogLocator

# Data:
#   lr_rf_1y_t   : float  — risk-free log return at time t (e.g., log(1.03) for 3%)
#   lr_t2        : float  — stock actual log return at time t2 = t + period_d
#   ema_var_d_t  : float  — daily variance as log(return)^2, current estimate at time t as EMA(span=365/3)
#   h_var_d      : float  — daily variance as log(return)^2, historical estimate over whole stock history

al6x

import './base'

p_([1, 2, 2].empty7())                       // => false

p_([[2], [1]].sort_())                       // => [[1], [2]]
p_([[1], [1]].uniq_())                       // => [[1]]
p_(equal7([1], [1]))                         // => true
p_([1, 2, 3].median_())                      // => 2
p_([1, 2, 3].min_())                         // => 1

al6x

import pymc as pm
import arviz as az

def fit_bayesian_slow(df, t):
  r_t = df['r_t'].values
  mvar_t_t0 = t * 0.69 * df['mvar_d_t0'].values
  r_rf_t_t0 = t * np.log(df['ar_rf_1y_t0'].values) / 365

  with pm.Model() as model:
    v0 = pm.HalfNormal('v0', sigma=1)

al6x

[0.07662950232169972, 0.08667983817520133, 0.08667983817520133, 0.08667983817520133, 0.18199001797952605, 0.2690013949691558, 0.1636408793113296, 0.18199001797952605, 0.18199001797952605, 0.07662950232169972, 0.07662950232169972, 0.18199001797952605, 0.18199001797952605, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.06833069950700488, 0.06833069950700488, 0.06833069950700488, 0.06833069950700488, 0.06833069950700488, 0.06833069950700488, 0.06833069950700488, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.07662950232169972, 0.08667983817520133, 0.06833069950700488, 0.1636408793113296, 0.1636408793113296, 0.1636408793113296, 0.1636408793113296, 0.2436835869848659, 0.2436835869848659, 0.

al6x

import numpy as np
import matplotlib.pyplot as plt
from sklearn.mixture import GaussianMixture
from scipy.stats import norm

def fit_normal_mixture(*, n_components, values, random_state, n_init):
    values = np.array(values).reshape(-1, 1)  # Convert to 2D array
    nmm = GaussianMixture(n_components, covariance_type='diag', random_state=random_state, n_init=n_init)
    nmm.fit(values)
    means = nmm.means_.flatten().tolist()

al6x

[3,2.9,2.9,3,3.1,3.2,3.1,3.4444,3.4444,3.4444,3.4444,3.2,3.7778,3.5556,3.3,3.5,3.7778,3.6667,3.4,3.3,3.3,3.2,3.3,3.3,3.4,3.3,3.4,3.4,3.4,3.0909,2.8333,3.0833,3.2727,3.2727,3.2727,3.3636,3.6364,3.5455,3.7273,4,3.9091,3.7273,3.8182,3.9091,3.7273,3.6364,3.5455,3.4545,3.5455,3.4545,3.4545,3.4545,3.6364,3.3333,3.25,3.25,3.25,3.25,3.1667,3.4545,3.4545,3.2727,3.0909,3.0909,3.0909,3.1818,3.2727,3.2727,3.2727,3.2727,3.2727,3.2727,3.2727,3.6,3.2727,3.1818,3.2727,3.1818,3.4,3.1,3.2,3.3,3.7778,3.4,3.6,3.6,3.5,3.6,3.5,3.4,3.1818,3.4,3.0909,3.1818,3.1818,3.4,3.4,3.2,3.3,3.4,3.5,3.6,3.6,3.5,3.5,3.6,3.2727,3.2727,3.2727,3.2727,3.7,3.7,3.9,4,4,4,4.2,3.9,4,4.4444,3.9,3.5455,3.5455,3.6364,4,4.1,4.1,4.1,4,4,3.7,4,4.2,4.3,4.3,4.4,4.5,4.6,4.8,5,5,4.6364,4.6364,4.5455,4.5455,4.1667,4.25,4.0833,4.0833,3.75,2.3846,2.5385,3,2.8462,2.6154,2.3846,2.2308,1.9286,2.5385,2.4286,2.5,2.3571,2.4286,2.3333,2.2,2.2,2.2,2.2667,2.3333,2.2667,2.2,2.1333,2.3571,2.1333,2.0625,1.8333,1.8333,1.7778,1.7222,1.8235,1.7647,1.8235,1.6471,1.7059,1.7647,1.941

al6x

import './mfun'

declare global {
  function fsize<T>(a: T[]): number
  function fempty7<T>(a: T[]): boolean
}

mfun(Array, function fsize<T>(a: T[]): number { return a.length })

mfun(Array, function fempty7<T>(a: T[]): boolean { return a.length == 0 })

al6x

table = [
  sort(
    map(
      (c) -> (type = :Without, moneyness = c.moneyness),
      filter((c) -> c.tenor == 856, normalized_jl)
    )
  )...,
  sort(
    map(
      (c) -> (type = "With CPI", moneyness = c.moneyness),

al6x

export {}

type Req = { id?: string, target: string, args?: unknown[] }
type Res = { id?: string, is_error: true, error: string } | { id?: string, is_error: false, result: unknown }

const target_cache: { [name: string]: unknown } = {}
async function process({ id, target: path, args }: Req): Promise<Res> {
  if (!(path in target_cache)) {
    const [module_name, ...names] = path.split('.')
    if (names.length < 1) return { id, is_error: true, error: `Invalid path: ${path}` }

al6x

// Use 'Prettify Symbols Mode' VSCode Extension to display `foo7` as `foo?`
declare global {
  function p(...args: unknown[]): void
  function inspect(v: unknown): void
  /** Converts anytihng to compact format `{ a: 1, b: [1, 'c']}`,
    * could be parsed with the `from_s`, relaxed json or yaml. */
  function to_s(v: unknown): string
  /** Parses both json, relaxed json, and output of the `to_s`, example `{ k: 1 }` */
  function from_s<T = unknown>(v: string): T
  function equal7<T>(a: T, b: T): boolean