m-dadej / downloading-and-aggregating-stocks Goto Github PK

There are some "minor" problems with data from bossa.

1. Some stocks from main platform are included in NC file and reverse
2. BNPPPL is full of NAs. There might be more such stocks
3. (my fault) there are still some funds in final dataset

Issue #1: tl;dr: changes in websites databases might unable reproducibilty of code (e.g. delisting stocks further giving NA in script). When downloading stocks from bossa or stooq, there are obviously changes in their stock database. There might be an issue when some stock is delisted, hence the number of stock changes and it might influence the repdorucibility of the code. Its because, i used to be a messy programmer, when i was making this script. Example that happen recently (10-06-2019). Stock with the name "SANPL2" was delisted because it is a right to a "SANPL" stoc, with the expiration day (7-6-2019). "SANPL2" was included in a vector of unwanted assets and when it wasnt found in the dataset of every stock, the script gave an NA in the vector of rows of unwanted stock tickers, making further part of code unable to be continued.

some parts of scripts might be missing after i divide the main script into 10 different ones