Desert Island Discs 1942-2020

This dataset contains details of all episodes of the BBC radio programme Desert Island Discs from its inception until early 2020. It was created for this article - https://www.musichistorystats.com/desert-island-discs/ - which explains how it was created and some of its limitations.

The Excel file contains two sheets, which were originally separate CSV files, and you might prefer to extract and work with them as such.

DID_castaways contains one row per episode (up to early 2020), with the following fields...
"episode_ref" - an episode identifier;
"date" - date of broadcast in YYYY-MM-DD format;
"name" - the name of the castaway as listed;
"std_name" - a standardised format of the castaway name, should be the same for those who appear more than once;
"gender" - gender of castaway (not 100% reliable - based on titles and first names);
"favTrack" - name of favourite track;
"luxury" - luxury item taken to the island;
"book" - book taken to the island;
"link" - URL of BBC website page for this episode;
"castaway_ref" - identifier for castaway (based on std_name).

DID_discs contains one row per disc, with the following fields...
"episode_ref" - as above;
"track_name" - track as listed;
"trackNo" - number of track in the episode (normally 1 to 8);
"favourite" - logical value indicating favourite track;
"std_artist" - as above;
"std_disc" - a standardised format version of the track name in the format "artist # title" (not 100% reliable - see article above for some of the reasons why not);
"disc_ref" - an identifier for the disc (based on std_disc);
"date" - the date of the episode;
"std_name" - standardised castaway name as above;
"gender" - as above;
"castaway_ref" - as above.

Missing data are generally indicated by "NA".

The file is encoded as UTF-8 rather than the Excel default. This should hopefully ensure that foreign characters are retained on different platforms.