Review of prototype 1 • financialDataAnalysis

The ‘financialDataAnalysis’ project was split up into 4 prototypes. This is a demonstration of of prototype 1.

library(financialDataAnalysis)

Prototype 1 includes the following features:

A large and rich dataset about a set of stocks, which can be freely used and analysed.
Two models, fit on a large dataset of historic stock prices, allowing you to predict the price of a variety of stocks.

The stock data contains 497 rows and 100 columns. Each row represents a particular stock.

The models can be used in the following way:

Note that the shown method for generating predictions will no longer work as of prototype 4. Instead, use the predict_price() function.

library(generics)
library(stats)
library(tibble)
library(bonsai)
library(lubridate)

First we need to create a data frame to predict on. Lets use make monthly predictions, going 6 months ahead.

data <- tibble(
  ticker = "GOOGL",
  ref_date = seq(
    from = today(),
    to = today() + months(6),
    by = "month"
  )
)
data
#> # A tibble: 7 × 2
#>   ticker ref_date  
#>   <chr>  <date>    
#> 1 GOOGL  2023-02-19
#> 2 GOOGL  2023-03-19
#> 3 GOOGL  2023-04-19
#> 4 GOOGL  2023-05-19
#> 5 GOOGL  2023-06-19
#> 6 GOOGL  2023-07-19
#> 7 GOOGL  2023-08-19

Now lets make predictions on the data using the monthly model.

predict(monthly_stock_model, data)

You can also add the predictions automatically to the data you have created.

augment(monthly_stock_model, data)