MISdata

End-to-end pipeline for stock market data analysis in R.

Installation · Examples · Function Reference · User Manual

Overview

MISdata provides a complete workflow for retrieving, cleaning, transforming, and analyzing stock market data. The package fetches index constituents from Wikipedia and OHLCV history from Yahoo Finance, then offers utilities for NA handling, time-series format conversion (via tsbox), period aggregation with OHLCV-aware semantics, exploratory analysis (line plots, ACF/PACF, STL decomposition), and ARIMA-based forecasting with multi-level confidence bands.

This package was developed as part of an undergraduate project and follows CRAN-compatible standards.

Workflow

flowchart LR
    A["Choose symbols<br/>get_index_components()<br/>sample_symbols()"]
    B["Fetch OHLCV data<br/>get_stock()"]
    C["Clean missing values<br/>clean_stock()"]
    D["Convert format<br/>convert_stock()"]
    E["Aggregate period<br/>change_period()"]
    F["Explore data<br/>plot_stock()<br/>get_acf() / get_pacf()<br/>plot_seasonal()"]
    G["Forecast series<br/>forecast_stock()"]

    A --> B --> C
    C --> D
    C --> E
    D --> F
    E --> F
    C --> F
    F --> G

Loading

Installation

Install the latest stable version from GitHub:

# install.packages("remotes")
remotes::install_github("MISDataGit/MISdata")

Examples

The examples below use Apple and Microsoft closing prices from January 1, 2023 through December 31, 2024.

Retrieve and visualize stock data

library(MISdata)

symbols <- c("AAPL", "MSFT")

stocks <- get_stock(
  symbols = symbols,
  start = "2023-01-01",
  end = "2024-12-31",
  columns = "Close"
)
stocks <- clean_stock(stocks, na_method = "trim")

history_plot <- plot_stock(
  stocks,
  symbols = symbols,
  column = "Close",
  title = "Daily Closing Prices: Apple and Microsoft"
)
history_plot

Forecast a stock series

apple_forecast <- forecast_stock(
  stocks,
  symbol = "AAPL",
  column = "Close",
  horizon = 30,
  ci_levels = c(80, 95)
)

summary(apple_forecast$model)
apple_forecast$plot

The forecast chart includes the fitted ARIMA model, a 30-step point forecast, and 80% and 95% confidence intervals.

To regenerate both README images from the package source:

Rscript tools/generate-readme-images.R

Function Reference

The package is organized into six modules:

Module	Functions	Purpose
fetch.R	get_index_components(), sample_symbols(), get_stock()	Retrieve index constituents and OHLCV data
clean.R	clean_stock()	Group-wise NA handling on long data frames
convert.R	convert_stock()	Convert between long df, xts, tsibble, zoo, ts
period.R	change_period()	OHLCV-aware period aggregation
eda.R	plot_stock(), get_acf(), get_pacf(), plot_seasonal()	Visualization and autocorrelation analysis
forecast.R	forecast_stock()	ARIMA forecasting via forecast::auto.arima()

For complete usage instructions, read the MISdata User Manual, or install the package and run ?MISdata or ?<function_name>.

Authors

• Batuhan Karaköy
• Enes Türkoğlu
• Metehan Kaygısız

License

GPL (>= 3). See LICENSE.md for the full text.