Icon

Challenge 11 Let's Dance1

<p><strong>Challenge 11: Let's Dance</strong></p><p><strong>Level:</strong> Medium<br><br><strong>Description:</strong> As a data-driven DJ, you are tasked with curating the perfect playlist to keep the crowd dancing during a two-hour event focused on Indian music. The datasets you are using comprise songs from 15 Indian languages, giving you a diverse range of tracks to work with. Your goal is to select songs with the highest danceability scores, ensuring that each track contributes to an energetic atmosphere throughout the event. You will use a dataset to choose the best tracks, sort them by danceability, calculate cumulative durations, and filter the playlist to stay within the two-hour limit.</p>
Challenge 11: Let's Dance

Level: Medium

Description: As a data-driven DJ, you are tasked with curating the perfect playlist to keep the crowd dancing during a two-hour event focused on Indian music. The datasets you are using comprise songs from 15 Indian languages, giving you a diverse range of tracks to work with. Your goal is to select songs with the highest danceability scores, ensuring that each track contributes to an energetic atmosphere throughout the event. You will use a dataset to choose the best tracks, sort them by danceability, calculate cumulative durations, and filter the playlist to stay within the two-hour limit.

Challenge 11: Let's Dance

Solution reasoning


  1. Upload via csv (poiting at the whole folder with 1 node)

  2. converting string to date and duration (with intermediate steps)

  3. adding duration by seconds for furthers calculations

  4. filtering out danceability = 0 and missing valueas ( circa 0,5%)

Interesting fact: there are an interesting number of same duplicated songs but with differente features (different danceability, maybe a remix but generally with same singer, duration). I retained only the song with the highest danceability score (Duplicate Row Filter). About 3640 songs purged (circa 10% original dataset).


VARIANT

Added a calculated custom_score variable:

  • 0,5*danceability

  • 0,2*energy

  • 0,2*valence

  • 0.1*popularity

Raw Solution



Variant with Custom Score



Variant with at least 1 song per language


Applied a rule to have at least 1 song per language

I hope to have realized good visuals

uload
CSV Reader
Data Viz
Normalizer
Column Renamer
Math Formula
Reference Row Filter
data cleaning
Normalizer
Normalizer
Loop End
Row Filter
Concatenate
filtering at leat 1 song per language
Group Loop Start
Moving Aggregator
descending orderdanceability
Sorter
descending ordercustom score
Sorter
Moving Aggregator
Moving Aggregator
Row Filter
Row Filter
Row Filter

Nodes

Extensions

Links