jonlachmann

library(lubridate)

# Function to count days, weekdays and weeks in a given month
counts <- function (year, month) {
  # Get the number of days in the month
  num_days <- days_in_month(as.Date(paste(year, month, "01", sep = "-")))

  # Initialize a counter
  weekday_count <- 0
  week_count <- 0

jonlachmann

# Generate mock data (two variable which go from 1-100 and 101-200 with some random noise).
data <- matrix(rnorm(200, 1:200, 0.5), 100, 2)

K <- ncol(data) # Number of variables in model
horizon <- 12 # Forecast horizon
p <- 2 # Lag order of model

# Create the lagged matrix of data X
X <- embed(data, p+1)[,-seq_len(K)]
X <- cbind(X, 1)

jonlachmann

from numpy import array
from numpy import linspace
from numpy import random
from numpy import zeros
from numpy import vstack
import torch


# Split a multivariate sequence into samples
def split_sequences(sequences, n_steps):

jonlachmann

esphome:
  name: nedis-16a

esp8266:
  board: esp01_1m

# Enable logging
logger:

# Enable Home Assistant API

jonlachmann

To be able to run address sanitizing in your own Rcpp/RcppEigen code, you need a version of R that is compiled with support for it. You also need to compile all Rcpp libraries that you will use using address sanitizing. Finally, your own code must also be compiled with it. This is the steps that I have found that works.

1. Download Docker for your platform

2. Install the rocker docker image and give it 8GB of memory (installation of Rcpp will fail on 2GB)
docker pull rocker/r-devel-san
docker container run --memory=8g -it --entrypoint bash rocker/r-devel-san

3. Set the Makevars to be used for ALL compiling of C++ code in R by creating /root/.R/Makevars with
PKG_CXXFLAGS = -fsanitize=address -I../inst/include

jonlachmann

# Title     : Multivariate Metropolis Hastings for covariance matrix estimation
# Objective : Estimate distribution of multivariate covariances
# Created by: jonlachmann
# Created on: 2020-05-17

# Jeffreys prior, log transformed
jeffreys_log = function (sigma) {
	-(nrow(sigma)+1)/2 * log(det(sigma))
}