Python DJ Playlist program

[Agena_]

Figure 4 shows selection of songs paul has store

 

 

rules:

1. duration of playlist is to be generated is entered in minutes

2. 25% of playlist left empty so paul can add songs that customers request

3. a playlist is then generated for the remaining duration using these rules:

            45% must be 5 star

            30% must be 4 star

            25% must be 3 star

4. songs with the greatest 'last selected' value should always be selected first

5. when a song is added to the playlist then the 'last selected' is set to 0

 

NOTE: 1 and 2 stars are not added to playlist. they are stored in case a customer requests during an event

[Agena_]

Plus im not that fluent so no hard stuff please

[minibois]

So... You want us to create your homework?

What have you finished about this assignment yet? What issue(s) are you running into?

[vorticalbox]

1 minute ago, Agena_ said:

Plus im not that fluent so no hard stuff please

So you want us to do you homework for you? I am going to do because it looks fun though it'll be tomorrow when I do but you should really try yourself and we can help with any problems you in counter

[vorticalbox]

So thought i would have play with this and this is what i came up with but before you look please please actually have a go your self you will learn nothing if you never try to solve the puzzle after you are done come back here and see how mine differs from yours.

 

output
 

# Data after the list generation to show the last was updated
                   name           artist  duration  rating  last
0       Jagged Boarders        Rob Smith       204       3     0
1           Last Minute         Sharpiee       165       3    19
2   It's gonna be great            Flybe       258       5     5
3                 Chaos  Trouble dodgers       302       5     0
4              Dreamin'        Rob Smith       214       3    16
5             Rock this        Ian Green       232       2    25
6           Let's Party            FlyBy       178       4     0
7           Lovin' life      Andrew ford       257       5     1
8                 Smile            Crazy       249       3    17
9          Keep It Real    Story Tellers       228       4     0
10                 Time          Various       232       5     4
11      Let's selebrate       Harry Bunn       249       5     0

# this is the play list that is also shuffled
              name           artist  duration  rating  last
0     Keep It Real    Story Tellers       228       4    11
1  Let's selebrate       Harry Bunn       249       5     7
2            Chaos  Trouble dodgers       302       5     9
3  Jagged Boarders        Rob Smith       204       3    20
4      Let's Party            FlyBy       178       4    13

# print the duration
Playlist length 19.35 mintues

 

On 9/27/2020 at 10:34 AM, Agena_ said:

Plus im not that fluent so no hard stuff please

 

incoming hard stuff

Spoiler

 

 

 

Spoiler

import pandas as pd
from math import floor

columns=['name','artist', 'duration', 'rating', 'last']

data = [
  ['Jagged Boarders', 'Rob Smith', 204, 3, 20],
  ['Last Minute', 'Sharpiee', 165, 3, 19],
  ["It's gonna be great", 'Flybe', 258, 5, 5],
  ['Chaos', 'Trouble dodgers', 302, 5, 9],
  ["Dreamin'", 'Rob Smith', 214,3,16],
  ['Rock this', 'Ian Green', 232, 2, 25],
  ["Let's Party", 'FlyBy', 178, 4,13],
  ["Lovin' life", 'Andrew ford', 257, 5,1],
  ['Smile', 'Crazy', 249, 3, 17],
  ['Keep It Real', 'Story Tellers',228,4,11],
  ['Time', 'Various', 232, 5, 4],
  ["Let's selebrate", 'Harry Bunn', 249, 5,7],
  ]

df = pd.DataFrame(data, columns=columns)


def get_songs(rating: int, duration: int) -> list:
  # get the songs
  song_data = df.loc[df['rating'] == rating].sort_values(by='last', ascending=False).values.tolist()
  songs = list()
  # turn mintues into seconds
  max_duration = duration * 60
  length = 0
  for song in song_data:
   name = song[0]
   song_duration = song[2]
   if length + song_duration > max_duration: break;
   index = df.index[df['name'] == name].tolist()[0]
   df.at[index, 'last'] = 0
   songs.append(song)
   length = length + song_duration
  return songs

def append_to_playlist(playlist, songs):
  temp = pd.DataFrame(songs, columns=columns)
  return playlist.append(temp, ignore_index=True)


length = 30 # 30 min playlist

s5Total = floor(length * 0.45)
s4Total = floor(length * 0.3)
s3Total = floor(length * 0.2)

# start the playlist with 5 star songs
playlist = pd.DataFrame(get_songs(5, s5Total), columns=columns)
# # add 4 star songs
playlist = append_to_playlist(playlist, get_songs(4, s4Total))
# # add 3 star songs
playlist = append_to_playlist(playlist, get_songs(3, s3Total))
# # shuffle that list 
playlist = playlist.sample(frac=1).reset_index(drop=True)

print(df) # song database

print(playlist) # our playlist
print(f'Playlist length {playlist["duration"].sum()/60} mintues')

 

 

[mariushm]

The problem is quite interesting, because of this:

 

* you have a total playlist time (60 minutes for example) and have to reserve 25% of that, so your actual time is 75% of input number.

* you then have 3 "pools" of audio tracks (45%, 30% , 25%) in which you have random audio tracks.

Then in theory you have to combine the 3 pools and sort by last used

 

Here's my php solution ... it does everything but the final sorting by last used... that's trivial :

 

 

<?php

$duration = isset($_GET['time'])? intval($_GET['time']) * 60 : 0;
$action = isset($_GET['action'])? $_GET['action'] : '';

// reset the database, generate random scores and last used for each track
if ($action=='reset') {

	$entries = [];
	$lines = explode("\r\n",file_get_contents('database.txt'));

	foreach ($lines as $line) {
		$l = explode("\t",$line);
		$t = explode(":",$l[3]);
		$timer = intval($t[0]) * 60 + intval($t[1]);
		$l[0] = mb_convert_encoding($l[0],'UTF-8','ASCII');
		$l[2] = mb_convert_encoding($l[2],'UTF-8','ASCII');
		
		array_push($entries, array('artist'=>$l[0],'track'=>$l[2],'time'=>$timer,'score'=>mt_rand(1,5),'last'=>mt_rand(0,60)));
	}

	$js = json_encode($entries,JSON_PRETTY_PRINT);
	file_put_contents( __DIR__ . '/database.json',$js);
	die('Playlist reset.');
}

// load database

$entries = json_decode(file_get_contents(__DIR__.'/database.json'),TRUE);
if (count($entries)<1) die('Database empty. Maybe reset it?');
foreach ($entries as $idx=>$entry) {
	$entries[$idx]['used'] = false;
}

// pick a track with desired score, suitable length and not selected before 
function find_track($len,$score) {
	global $entries;
	$ids = [];
	foreach ($entries as $i => $e) {
		if (($e['time']<=$len) && ($e['score']==$score) && ($e['used']==FALSE)) {
			array_push($ids,$i);
		}
	}
	if (count($ids)==0) return FALSE;
	$result = mt_rand(0,count($ids)-1);
	return $ids[$result];
}

if ($duration > 0) {
	$time_total = round($duration * 0.75,0);
	$pools = array();
	$pools[0] = array('score'=>5, 'duration'=> intval(0.45 * $time_total) , 'length'=>0, 'tracks'=> []);
	$pools[1] = array('score'=>4, 'duration'=> intval(0.30 * $time_total) , 'length'=>0, 'tracks'=> []);
	$pools[2] = array('score'=>3, 'duration'=> intval(0.25 * $time_total) , 'length'=>0, 'tracks'=> []);
	
	for ($i=0;$i<3;$i++) {
		$seconds = $pools[$i]['duration'];
		$track = TRUE;
		while ($track = find_track($seconds,$pools[$i]['score'])) {
				array_push($pools[$i]['tracks'],$entries[$track]); // copy track to pool track array
				$pools[$i]['length'] += $entries[$track]['time'];  // just for stats
				$seconds = $seconds - $entries[$track]['time'];
				$entries[$track]['used'] = TRUE; // so it won't be picked twice
		}
	}
	// print the tracks
	for ($i=0;$i<3;$i++) {
		echo '<p>Pool '.$i.': max='.$pools[$i]['duration'].', used='.$pools[$i]['length'].', tracks='.count($pools[$i]['tracks']).'</p>';
		echo '<table>';
		echo '<tr><td>Artist</td><td>Track</td><td>Time</td><td>Score</td><td>Used last</td></tr>';
		foreach ($pools[$i]['tracks'] as $track) {
			echo '<tr><td>'.$track['artist'].'</td><td>'.$track['track'].'</td><td>'.
							$track['time'].'</td><td>'.$track['score'].'</td><td>'.$track['last'].'</td></tr>';
		}
		echo '</table>';
		
	}
	
die('Done.');
}


die('Usage:<br/>
?action=reset  - reset database<br/>
?time=[minutes] - make playlist with [minutes] minutes<br/>
');
  	
?>

 

example output:

 

 

The source code and example database is attached.

 

 

playlist.zip

[vorticalbox]

12 hours ago, mariushm said:

Then in theory you have to combine the 3 pools and sort by last used

Quote

4. songs with the greatest 'last selected' value should always be selected first

You should be selecting your pools by last selected.

[mariushm]

In that case, all that needs to be done is modify the find_track function to sort the suitable tracks by last used, then always return the first track (with most used value) instead of a random one. Probably 2 extra lines, and one line modified... easy.

[vorticalbox]

1 hour ago, mariushm said:

In that case, all that needs to be done is modify the find_track function to sort the suitable tracks by last used, then always return the first track (with most used value) instead of a random one. Probably 2 extra lines, and one line modified... easy.

yeah its actually a fun little project, which my college classes had problems like this