In ilangurudev/IndianStocksR: End of Day Data Downloader and Updater for Indian Stock Markets

pacman::p_load(knitr,
               kableExtra)



make_table <- function(tbl) {
  # DT::datatable(tbl)
  knitr::kable(tbl, "html") %>%
        kable_styling(bootstrap_options =
                        c("striped", "hover", "responsive"),
                      full_width = FALSE,
                      position = "left") %>% 
    scroll_box(width = "1000px", height = "500px")

}

Introduction

The IndianStocksR package is used to download the end of day data of all stocks in the two primary Indian stock markets, NSE and BSE. The end of data data is provided free by the two stock exchanges from their websites and consists of information like the open, high, low, close among others for each script that's traded in them.

The data can be accessed from their websites based on the date formatted in a certain way. The R-Bloggers article was the source of inspiration for the package. However, the package modularizes the code, tweaks a lot of things and creates a much more accessible API that's more powerful in the sense that it abstracts away the complexity from the user.

It is advised to create a folder and set the working directory to that folder before we start work. Even better, if you're working from R studio is to create an R project for downloading the data and working on your analysis.

The package is currently getting submitted to CRAN after which a simple install.packages("IndianStocksR") will get it installed. But for now, it is available on github.

install.packages("devtools")
devtools::install_github("ilangurudev/IndianStocksR")

After installation, we load the package. The package basically creates data frames and hence plays along well with the concepts of tidy data and the tidyverse. So it is highly encouraged to load that package as well

library(IndianStocksR)
library(tidyverse)

Main Functions

download_stocks

The workhorse of the package is the function download_stocks. However, you will rarely have to use it. It still pays to understand the parameters as it is the basis of the other functions that you will probably use.

download_stocks(date = "2018-07-20", exchange = c("nse", "bse"), dest_path = "./data", quiet = FALSE)

• The date parameter can be a date object (and defaults to today). It can also be a string (yyyy-mm-dd) or a number that can be parsed as a date by lubridate::as_date(). For instance, "2018-05-21" is a valid date.
• The exchange can either be "nse" or "bse".
• The dest_path specifies where you want the data files to get downloaded. It defaults to the data folder in the current working directory (which it will create if not found). This is why it is advisable to work in a project. This keeps all the data files of a project organized. If the path you specify is not found, an error is thrown.
• The quiet parameter controls whether you want the download status messages or not.

The main purpose of this function is to download data from the specified exchange on the mentioned date. If data is not available for the date you specified, you will get an error.

download_stocks_period

The function you'll probably have to use first is the download_stocks_period

df_period <- 
  download_stocks_period(start = "2018-07-21",
                         end = "2018-07-26",
                         exchange = c("both", "nse", "bse"),
                         dest_path = "./data",
                         compile = TRUE,
                         delete_component_files = TRUE,
                         quiet = FALSE)

• start and end: The download stocks period downloads data for all the dates in the date range specified by start and end. start defaults to today - 8 days and end defaults to today. If today is r lubridate::today(), then end takes that value and start takes the value r lubridate::today()-8. However, it makes sense to make start today - 365 or specify the actual date from when you want the data. You could change the end value too if you want data for a specific date range. The start and end values follow the same rules as the date parameter in download_stocks
• The exchangefunction's behavior is pretty straightforward. Downloads data for the date range from NSE if "nse" or BSE from "bse" or both NSE and BSE if "both". Defaults to "both"
• The dest_path does the same job as it does in download_stocks
• The compile parameter compiles all the downloaded files into one file (if exchange is "both", one compiled file for "nse", one for "bse" and one combined). This option is by default on as compiled files are much more tractable for analysis.
• The delete_component_files deletes everything apart from the compiled files. This keeps the work space clean and more efficient for updating.
• The quiet does the same job as it does in download_stocks

Let's take a look at df_period.

df_period %>% slice(1:200)

df_period %>% slice(1:15) %>% make_table()

The function returns the compiled files apart from writing them out as a csv.

update_stocks

Once you have the download_stocks_period run, you can update the database later by running the update_stocks

df_updated <- 
  update_stocks(data_path = "./data",
                till = lubridate::today(),
                exchange = c("both", "nse", "bse"),
                compile = TRUE,
                delete_component_files = TRUE)

Most of this parameters have been discussed before. This function scans all the files in the directory and finds out the date till which there is data and downloads data from the day after till the date mentioned by till. If there are no files inside the specified folder, it downloads data from today - 8 till the date mentioned by till. You rarely have to tweak the till function. It's primarily used to update till the current day.

Let's take a look at df_updated.

df_updated %>% slice(1:200)

df_updated %>% slice(1:15) %>% make_table()

A note on defaults

Except the date parameters, one rarely has to tweak the defaults. The defaults are designed to work optimally.

Bugs, comments, suggestions and feature requests

This is just an initial version of the package and I expect to see a few bugs. I'd be very happy if you create github issues if you run into anything. Suggestions and feature requests welcome. Feel free to comment what you think of the package.

Thanks for reading! Cheers!

ilangurudev/IndianStocksR documentation built on May 12, 2019, 4:36 a.m.