Steps for formatting Date and Creating Index

Code samples

Code 
#source("common_functions.R") # should i be using this one?? not a good idea since many of the formulas return different values

# Loading R packages. originally was using common functions but trying not to use
if (!require("pacman")) install.packages("pacman")
pacman::p_load(tidyverse,
               tsibble, fable,
               feasts, tsibbledata,
               fable.prophet,
               patchwork,
               lubridate,
               rio,
               ggplot2,
               kableExtra, tidyquant
               )

The following is steps to check in what format the date column is in in a new dataset. Then to convert to DATE format to do time series research

Code 
# the rename is for a different df

# Step 1: Import your data
df <- read.csv("https://byuistats.github.io/timeseries/data/Vessels_Trade_US.csv") 
# We'll assume the data has a column called 'date' (replace with the actual column name) and a 'value' column


# Step 2: Check the structure of the 'date' column to verify its type
str(df$date)  # This will show you if the date column is a character, Date, or something else


# Convert 'date' column to Date type if it's in character format 
# 2.1 reomve unwanted columns or
df1 <- df |>
  mutate(date = lubridate::mdy(date) # date is date columm name
         # mdy(date) mdy is current order of date, lubridate will format to ymd. 
       #dplyr::select(-comments) # remove unwanted columns
         ) |>
  rename(x = constructionequip_ord, y = constructionequip_ship) # rename columns

# 2.1 Convert 'Date' column to yearquarter format
df$Date <- lubridate::ymd(df$Date) # df2



# Verify the 'date' column is now in the correct format (should be Date type)
str(df$date)  # Should now return 'date'

Doing mutate by getting weekly average

Code 
# Set symbol and date range for Apple
symbol <- "AAPL"
date_start <- "2022-01-01"
date_end <- "2024-01-01"

# Download the stock data
stock_df <- tq_get(symbol, from = date_start, to = date_end, get = "stock.prices")

# Convert to a tsibble
stock_ts <- stock_df %>%
  mutate(dates = lubridate::ymd(date), value = adjusted) %>%
  mutate(year_week = yearweek(dates)) |>
  group_by(year_week) |>
  summarise(value = mean(value)) |>
  ungroup() |>
  as_tsibble(index = year_week)

# Time plot of the daily closing prices
autoplot(stock_ts, value) +
  labs(title = "Time Plot of Apple (AAPL) Daily Closing Prices",
       x = "Date", y = "Closing Price (USD)")

the code below was taken from project one, made to plot the time series without doing the monthly mean. The first two lines of code are missing

data is for daily data.

Code 
# this code is also not complete


# the first 2 ts can almost be taken from the previous code.
# this code was first use but replace by the above code

  dplyr::select(dates, year, months, value)  |> # ts3
  arrange(dates) |> # ts4
  mutate(index = tsibble::yearmonth(dates)) |> # ts5
  as_tsibble(index = index) |> # ts6
  dplyr::select(index, dates, year, months, value) |> # ts7
  rename(Vessels = value) # rename value to emphasize data context
vessels_ts |> #ts8
  autoplot(.vars = Vessels) +
  labs(
    x = "Month",
    y = "Vessels Cleared Monthly",
    title = "Vessels Cleared Monthly in Foreing Trade for United Sates"
  ) +
  theme_minimal() +
  theme(plot.title = element_text(hjust = 0.5))

working with two variables and period column

Code 
# I think this file is all for samples, and I edit them so now the code is all mix up hence all the eval=false! sucks 

df <- rio::import("https://byuistats.github.io/timeseries/data/constructionequip_manu_orders_shipments.csv")

# str(df$date)


df1 <- df |>
  mutate(date = lubridate::mdy(date),
         constructionequip_ord = as.numeric(constructionequip_ord), # make sure numeric for x variables
         constructionequip_ship = as.numeric(constructionequip_ship)
         ) |>
  rename(x = constructionequip_ord, y = constructionequip_ship) |> # renames columns and converts to numeric
  select(date, x, y) # re orders and or removes not selected columns
  
df2 <- df1 |> # this makes a new df so either df before or this one is use. 
  mutate(obs = row_number()) |> # makes new column with periods
  select(obs, x, y)

# can the obs code to make multiple df with different columns.

data has gaps

Tyson notes

  1. filling with previous variable
    1. Then there is 100% correlation with the previous variable
  2. Filling with average
    1. some data’s average, like weather can vary so it can trow off random,
    2. using the same variable as last year is an option, it will just mess with the seasonality
  3. taking the sum of the lag and lead periods, and divide by two to replace missing
    1. its a good one
Code 
df0 <- rio::import("https://byuistats.github.io/timeseries/data/UNRATENSA.csv") |>
    mutate(YM = yearmonth(lubridate::mdy(date)))
df <- as_tsibble(df0, index = YM) |>
  select(ym, cdebt)
interval(df) # gives interval: M, D ot Y etc
has_gaps(df) # false if none and vice versa True

Reading different file formats

Code 
# wine_dat <- read_table("data/wine.dat") # resource 3.4.2
Back to top