Week3
2023-09-10
R Markdown
This is an R Markdown document. Markdown is a simple formatting syntax for authoring HTML, PDF, and MS Word documents. For more details on using R Markdown see http://rmarkdown.rstudio.com.
When you click the Knit button a document will be generated that includes both content as well as the output of any embedded R code chunks within the document. You can embed an R code chunk like this:
##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionNote that the echo = FALSE parameter was added to the
code chunk to prevent printing of the R code that generated the
plot.
data1<-head(data,40)Grouping Data by 1
data_group1 <- data1 %>%
group_by(artistName) %>%
summarise(avg_msplayed = mean(msPlayed),
Maximum = max(msPlayed))
print(data_group1)## # A tibble: 36 × 3
## artistName avg_msplayed Maximum
## <chr> <dbl> <int>
## 1 AURORA 1945555 1945555
## 2 Austyn Johnson 60453 60453
## 3 Avenged Sevenfold 40539 40539
## 4 Beyoncé 233 233
## 5 Bhikhari Bala 57728 57728
## 6 Black Hill 229535 229535
## 7 Brent Faiyaz 104821 104821
## 8 Catie Turner 6158627 6158627
## 9 Chris Lake 192455 192455
## 10 Clean Bandit 246584 246584
## # ℹ 26 more rowssummary(data_group1)## artistName avg_msplayed Maximum
## Length:36 Min. : 233 Min. : 233
## Class :character 1st Qu.: 116204 1st Qu.: 116204
## Mode :character Median : 217668 Median : 217668
## Mean :1184368 Mean :1229712
## 3rd Qu.:1297857 3rd Qu.:1701588
## Max. :9160154 Max. :9160154ggplot(data_group1, aes(artistName,Maximum)) + geom_col(fill = "#bb0a69",colour = "yellow") + labs(title = "Avg msplayed by Artist", 'artistName','avg_maplayed')+theme(axis.text.x = element_text(angle = 90))
Calculating expected probability for each group based on ‘msplayed’
data_group1 <- data_group1 %>%
mutate(exp_prob = (avg_msplayed/ sum(avg_msplayed)))
print(data_group1)## # A tibble: 36 × 4
## artistName avg_msplayed Maximum exp_prob
## <chr> <dbl> <int> <dbl>
## 1 AURORA 1945555 1945555 0.0456
## 2 Austyn Johnson 60453 60453 0.00142
## 3 Avenged Sevenfold 40539 40539 0.000951
## 4 Beyoncé 233 233 0.00000546
## 5 Bhikhari Bala 57728 57728 0.00135
## 6 Black Hill 229535 229535 0.00538
## 7 Brent Faiyaz 104821 104821 0.00246
## 8 Catie Turner 6158627 6158627 0.144
## 9 Chris Lake 192455 192455 0.00451
## 10 Clean Bandit 246584 246584 0.00578
## # ℹ 26 more rowsggplot(data_group1, aes(artistName,exp_prob)) + geom_col(fill = "#15c8b1",colour = "red") + labs(title = "Avg msplayed by Artist", 'artistName','avg_maplayed')+theme(axis.text.x = element_text(angle = 90))
Assigning ‘anomaly’ tag to the lowest probability group
anomaly1 <- data_group1 %>%
filter(exp_prob == min(exp_prob)) %>%
pull(artistName)
print(anomaly1)## [1] "Beyoncé"Assigning the lowest probability group an “anomaly” tag and Translating anomaly back into your original data frame i.e data1
data1 <- data1 %>%
mutate(anomaly_1 = ifelse(artistName == anomaly1, "Anomaly", "Normal"))
print(data1)## trackName artistName msPlayed
## 1 A Better Place Project AER 119999
## 2 A Dangerous Thing AURORA 1945555
## 3 A Different Way (with Lauv) DJ Snake 66060
## 4 A Drug From God Chris Lake 192455
## 5 A Gift Of A Thistle James Horner 97568
## 6 A Little Bit Yours JP Saxe 99339
## 7 A Little More Catie Turner 6158627
## 8 A Little Piece of Heaven Avenged Sevenfold 40539
## 9 A Million Dreams Ziv Zaifman 269453
## 10 A Million Dreams (Reprise) Austyn Johnson 60453
## 11 A Moment Apart ODESZA 300828
## 12 A New Beginning Yasumu 143378
## 13 A Night to Remember High School Musical Cast 1514502
## 14 A Night to Remember - Original Version High School Musical Cast 832507
## 15 A Place Among the Stars Hans Zimmer 1670916
## 16 A Sky Full of Stars Taron Egerton 372839
## 17 A Soulmate Who Wasn’t Meant to Be Jess Benko 9160154
## 18 A wild river to take you home Black Hill 229535
## 19 A+E (feat. Kandaka Moore & Nikki Cislyn) Clean Bandit 246584
## 20 A-Punk Vampire Weekend 1804847
## 21 AGEN WIDA JOYRYDE 198997
## 22 ALIEN SUPERSTAR Beyoncé 233
## 23 ALL MINE Brent Faiyaz 104821
## 24 ALLERGIC clide 155200
## 25 AMAZING Rex Orange County 6890534
## 26 AMNESIA (feat. Boy In Space) DREAMDNVR 3734865
## 27 ANGOSTURA keshi 171018
## 28 ANKLES Jessie Reyez 187423
## 29 ANUBIS KUTE 262699
## 30 AVOID ME KUTE 800336
## 31 AVOID ME 3 KUTE 40239
## 32 Aaftaab The Local Train 2540803
## 33 Aag Lage Chahe Basti Mai Sirazee 27373
## 34 Aahe Neelasaila Bhikhari Bala 57728
## 35 Aaj Bhi Vishal Mishra 1793602
## 36 Aaj Bhi - WORMONO LoFi Remake Vishal Mishra 76002
## 37 Aaja We Mahiya Imran Khan 2400332
## 38 Aajkal Tere Mere Pyar Ke Charche Suman Kalyanpur 205800
## 39 Aankhein Khuli Jatin-Lalit 173287
## 40 Aap Ki Kashish Himesh Reshammiya 333653
## genre danceability energy key loudness speechiness
## 1 ambient guitar 0.4960 0.2550 9 -17.984 0.0283
## 2 art pop 0.5410 0.5560 11 -6.150 0.0356
## 3 edm 0.7840 0.7570 8 -3.912 0.0384
## 4 bass house 0.7140 0.8830 9 -4.430 0.0625
## 5 orchestral soundtrack 0.0828 0.0120 9 -36.045 0.0451
## 6 alt z 0.5980 0.2950 1 -8.553 0.0276
## 7 alt z 0.7920 0.4840 4 -9.897 0.1920
## 8 alternative metal 0.4860 0.8810 2 -5.623 0.0474
## 9 show tunes 0.2650 0.3120 7 -11.689 0.0569
## 10 show tunes 0.2530 0.1390 6 -17.067 0.0414
## 11 chillwave 0.5150 0.6630 7 -7.896 0.0328
## 12 lo-fi study 0.6950 0.2870 0 -14.608 0.0499
## 13 post-teen pop 0.7980 0.9040 0 -3.316 0.0476
## 14 post-teen pop 0.7980 0.9040 0 -3.316 0.0476
## 15 german soundtrack 0.1710 0.0325 4 -25.382 0.0430
## 16 hollywood 0.4400 0.6760 6 -7.570 0.0582
## 17 gen z singer-songwriter 0.5710 0.0274 9 -20.274 0.0649
## 18 instrumental post-rock 0.4580 0.3610 6 -12.066 0.0353
## 19 pop 0.8080 0.5190 5 -8.268 0.0403
## 20 baroque pop 0.5510 0.8190 2 -4.489 0.0525
## 21 bass house 0.7840 0.9130 1 -1.208 0.1770
## 22 pop 0.5450 0.6410 10 -6.398 0.0998
## 23 r&b 0.6170 0.3780 9 -8.540 0.0315
## 24 singer-songwriter pop 0.6090 0.4280 11 -7.796 0.0244
## 25 bedroom pop 0.7190 0.4610 8 -8.191 0.0454
## 26 dark r&b 0.7080 0.4200 9 -9.478 0.1210
## 27 chill r&b 0.5360 0.6590 6 -9.732 0.0483
## 28 canadian contemporary r&b 0.5140 0.7490 8 -4.336 0.4030
## 29 aggressive phonk 0.6270 0.9930 1 -1.304 0.0714
## 30 aggressive phonk 0.6560 0.9640 6 -2.087 0.0964
## 31 aggressive phonk 0.7060 0.9330 10 -0.096 0.0721
## 32 hindi indie 0.4640 0.3910 6 -9.884 0.0384
## 33 himachali pop 0.6130 0.5510 9 -8.613 0.1980
## 34 odia bhajan 0.3280 0.5710 1 -6.913 0.0395
## 35 desi pop 0.5810 0.3040 1 -9.707 0.0274
## 36 desi pop 0.3070 0.2650 1 -14.701 0.0497
## 37 desi hip hop 0.6390 0.5940 10 -6.497 0.0347
## 38 classic bollywood 0.5560 0.7920 9 -10.683 0.0715
## 39 afghan pop 0.7920 0.6350 7 -13.940 0.0863
## 40 filmi 0.6900 0.7800 8 -7.412 0.0489
## valence tempo id duration_ms anomaly_1
## 1 0.0809 141.961 2oC9Ah7npALCCPW5DC1gob 120000 Normal
## 2 0.1060 105.886 0PDlmmYkuQCUAFhMXvtlsU 215573 Normal
## 3 0.5870 104.996 1YMBg7rOjxzbya0fPOYfNX 198286 Normal
## 4 0.8190 126.016 4skbQNtyjy8A7mo8oqe2oD 192455 Normal
## 5 0.0578 169.895 3Yvi5NkUrSppVwrMHYkB6u 97667 Normal
## 6 0.3140 87.025 00cBcYOlnHoXX9ver3cmdE 225680 Normal
## 7 0.2450 96.939 2thRPgNgcWAwAYxP4HWdji 224528 Normal
## 8 0.6670 144.997 1BLfQ6dPXmuDrFmbdfW7Jl 480707 Normal
## 9 0.0998 54.747 0RoA7ObU6phWpqhlC9zH4Z 269453 Normal
## 10 0.1020 74.369 66y7x28jXOPrcmu3D5Zjh6 60453 Normal
## 11 0.0756 120.024 59wlTaYOL5tDUgXnbBQ3my 234244 Normal
## 12 0.3790 128.028 0xtbVIWkbfu5G6TgCVmvVn 143379 Normal
## 13 0.9660 135.008 3RuXekS5criwismaaj87iF 236400 Normal
## 14 0.9660 135.008 2BHu6HOFXgHR936cj10XIi 236400 Normal
## 15 0.0284 70.821 7KNoh3Bn4QfgkIGBywkAB0 207118 Normal
## 16 0.4290 125.013 2VeLdiYiILcCOEV0izfzEW 206438 Normal
## 17 0.0870 124.209 0gOz9JUXsaKVzLTSmFDtdo 316556 Normal
## 18 0.2310 86.966 0UFkbnxj34vZVgwwEDy29e 229536 Normal
## 19 0.3840 114.977 1Qi6CGoqTG1UvFcuRzKRNL 246584 Normal
## 20 0.8460 174.917 3AydAydLzyyZutA0375XIz 137760 Normal
## 21 0.5900 126.054 35V6qXUDCvwZdmbKe5PlPG 199048 Normal
## 22 0.4640 121.892 1Hohk6AufHZOrrhMXZppax 215460 Anomaly
## 23 0.2200 141.967 3XgGQ1wjo5khvq2UImjyNF 216063 Normal
## 24 0.6270 96.071 6DP8InyxdyYChHb2tcV6ia 155200 Normal
## 25 0.5480 123.915 3OM6qQmdFV6uy61GIqpRtf 209200 Normal
## 26 0.3170 145.981 1UyDQh3HBJwL6IxR9mXipZ 175827 Normal
## 27 0.1990 170.035 38umMmZQdeoOG7Zojor4g3 171020 Normal
## 28 0.4160 179.623 3ZWHzPucGUZvgNleqllzdK 169240 Normal
## 29 0.5800 109.975 1aqP9g3lLR8bpFtkCTm41N 131455 Normal
## 30 0.9650 114.007 5FX30idriKlEIRYTxpNf65 105789 Normal
## 31 0.7610 114.042 7dXyRgDZlQnA5OU1SPD1oe 120000 Normal
## 32 0.4180 106.887 4LtSTc3xANVhYeeN69nscM 233472 Normal
## 33 0.4820 110.243 0gbNOv3x5M87aczFOIEzfp 142410 Normal
## 34 0.6540 186.602 3fB2jhxOmwMsTxCYpjYjd4 313748 Normal
## 35 0.0973 109.006 65jpYG7TQCh0T7TtKO4Rjj 241206 Normal
## 36 0.0569 94.115 0muXl7VdnqjsMH3IDS2JFI 217976 Normal
## 37 0.3780 80.034 75Ak0cit8hV604Wjlq1gf2 232067 Normal
## 38 0.7760 98.554 2uiAZVVKqqYlM6KM9POTYq 309493 Normal
## 39 0.6480 122.004 16XEVyPh5NT31CAAqPbxQF 422713 Normal
## 40 0.7180 110.963 35Rrbt58ADLxyo1lu7xhZu 333653 Normal##Conclusion for the 1st “group by” data frame: ##
##Here we can see that Beyonce is having the less expected probability compared to others so Anomaly tag is attached to it after translating to original data.##
possible_combinations <- expand.grid(artistName = unique(data1$artistName),
msPlayed = unique(data1$msPlayed))
print(head(possible_combinations))## artistName msPlayed
## 1 Project AER 119999
## 2 AURORA 119999
## 3 DJ Snake 119999
## 4 Chris Lake 119999
## 5 James Horner 119999
## 6 JP Saxe 119999missing_combinations <- anti_join(possible_combinations, data1, by = c("artistName", "msPlayed"))
print(head(missing_combinations))## artistName msPlayed
## 1 AURORA 119999
## 2 DJ Snake 119999
## 3 Chris Lake 119999
## 4 James Horner 119999
## 5 JP Saxe 119999
## 6 Catie Turner 119999Calculating the frequency of each combination
data_for_combination<-head(data,8000)
combination_counts <- data_for_combination %>%
group_by(artistName, msPlayed) %>%
summarise(count = n(), .groups = 'drop')
print(combination_counts)## # A tibble: 4,288 × 3
## artistName msPlayed count
## <chr> <int> <int>
## 1 !!! 3130 1
## 2 !!! 227066 2
## 3 $NOT 360000 2
## 4 $uicideboy$ 120005 1
## 5 11:11 Music Group 134582 2
## 6 12 Stones 206562 1
## 7 1nonly 270757 2
## 8 2 Chainz 50807 1
## 9 22 Void Beats 828909 1
## 10 24Herbs 263 1
## # ℹ 4,278 more rowsAfter calculating the frequencies we got different counts for different artistNames so we can analyze which artists have how many combinations with msPlayed.
Understanding well about Combinations and Missing Combinations
combination_data <- table(data1$artistName, data1$msPlayed)
print(head(combination_data))##
## 233 27373 40239 40539 57728 60453 66060 76002 97568 99339
## AURORA 0 0 0 0 0 0 0 0 0 0
## Austyn Johnson 0 0 0 0 0 1 0 0 0 0
## Avenged Sevenfold 0 0 0 1 0 0 0 0 0 0
## Beyoncé 1 0 0 0 0 0 0 0 0 0
## Bhikhari Bala 0 0 0 0 1 0 0 0 0 0
## Black Hill 0 0 0 0 0 0 0 0 0 0
##
## 104821 119999 143378 155200 171018 173287 187423 192455
## AURORA 0 0 0 0 0 0 0 0
## Austyn Johnson 0 0 0 0 0 0 0 0
## Avenged Sevenfold 0 0 0 0 0 0 0 0
## Beyoncé 0 0 0 0 0 0 0 0
## Bhikhari Bala 0 0 0 0 0 0 0 0
## Black Hill 0 0 0 0 0 0 0 0
##
## 198997 205800 229535 246584 262699 269453 300828 333653
## AURORA 0 0 0 0 0 0 0 0
## Austyn Johnson 0 0 0 0 0 0 0 0
## Avenged Sevenfold 0 0 0 0 0 0 0 0
## Beyoncé 0 0 0 0 0 0 0 0
## Bhikhari Bala 0 0 0 0 0 0 0 0
## Black Hill 0 0 1 0 0 0 0 0
##
## 372839 800336 832507 1514502 1670916 1793602 1804847
## AURORA 0 0 0 0 0 0 0
## Austyn Johnson 0 0 0 0 0 0 0
## Avenged Sevenfold 0 0 0 0 0 0 0
## Beyoncé 0 0 0 0 0 0 0
## Bhikhari Bala 0 0 0 0 0 0 0
## Black Hill 0 0 0 0 0 0 0
##
## 1945555 2400332 2540803 3734865 6158627 6890534 9160154
## AURORA 1 0 0 0 0 0 0
## Austyn Johnson 0 0 0 0 0 0 0
## Avenged Sevenfold 0 0 0 0 0 0 0
## Beyoncé 0 0 0 0 0 0 0
## Bhikhari Bala 0 0 0 0 0 0 0
## Black Hill 0 0 0 0 0 0 0missing_combination <- which(combination_data == 0, arr.ind = TRUE)
print(head(missing_combination))## row col
## AURORA 1 1
## Austyn Johnson 2 1
## Avenged Sevenfold 3 1
## Bhikhari Bala 5 1
## Black Hill 6 1
## Brent Faiyaz 7 1Group by 2
data2<-head(data,40) data_group2 <- data %>%
group_by(genre) %>%
summarise(median_duration = median(duration_ms),
Minimun2 = min(duration_ms))
print(head(data_group2))## # A tibble: 6 × 3
## genre median_duration Minimun2
## <chr> <dbl> <int>
## 1 a cappella 269636 269636
## 2 abstract 465200 465200
## 3 abstract beats 73898 73898
## 4 abstract hip hop 283747 283747
## 5 acoustic opm 244800 244800
## 6 acoustic pop 232941 186584summary(data_group2)## genre median_duration Minimun2
## Length:523 Min. : 65800 Min. : 10027
## Class :character 1st Qu.:173634 1st Qu.:132486
## Mode :character Median :200673 Median :171467
## Mean :206649 Mean :177224
## 3rd Qu.:230168 3rd Qu.:211425
## Max. :621659 Max. :621659ggplot(data_group2, aes(genre,median_duration)) + geom_point(size=3.5, color="#ff9600") + labs(title = "Avg duration played according to genre", 'genre','avg_duration')+theme(axis.text.x = element_text(angle = 90))
ggplot(data_group2, aes(genre,median_duration)) +
geom_boxplot()+theme(axis.title.x=element_text(colour="blue"),axis.title.y = element_text(colour="blue"))+ labs(title = "genre vs median_duration")
data_group2 <- data_group2 %>%
mutate(exp_prob2 = (median_duration / sum(median_duration)))
print(head(data_group2))## # A tibble: 6 × 4
## genre median_duration Minimun2 exp_prob2
## <chr> <dbl> <int> <dbl>
## 1 a cappella 269636 269636 0.00249
## 2 abstract 465200 465200 0.00430
## 3 abstract beats 73898 73898 0.000684
## 4 abstract hip hop 283747 283747 0.00263
## 5 acoustic opm 244800 244800 0.00227
## 6 acoustic pop 232941 186584 0.00216anomaly2 <- data_group2 %>%
filter(exp_prob2 == min(exp_prob2)) %>%
pull(genre)
print(anomaly2)## [1] "ballet class"ggplot(data_group2, aes(genre, exp_prob2)) +
geom_point()+
geom_smooth()## `geom_smooth()` using method = 'loess' and formula = 'y ~ x'
data2 <- data %>%
mutate(anomaly_1 = ifelse(artistName == anomaly1, "Anomaly", "Normal"))
print(head(data2))## trackName artistName msPlayed genre
## 1 A Better Place Project AER 119999 ambient guitar
## 2 A Dangerous Thing AURORA 1945555 art pop
## 3 A Different Way (with Lauv) DJ Snake 66060 edm
## 4 A Drug From God Chris Lake 192455 bass house
## 5 A Gift Of A Thistle James Horner 97568 orchestral soundtrack
## 6 A Little Bit Yours JP Saxe 99339 alt z
## danceability energy key loudness speechiness valence tempo
## 1 0.4960 0.255 9 -17.984 0.0283 0.0809 141.961
## 2 0.5410 0.556 11 -6.150 0.0356 0.1060 105.886
## 3 0.7840 0.757 8 -3.912 0.0384 0.5870 104.996
## 4 0.7140 0.883 9 -4.430 0.0625 0.8190 126.016
## 5 0.0828 0.012 9 -36.045 0.0451 0.0578 169.895
## 6 0.5980 0.295 1 -8.553 0.0276 0.3140 87.025
## id duration_ms anomaly_1
## 1 2oC9Ah7npALCCPW5DC1gob 120000 Normal
## 2 0PDlmmYkuQCUAFhMXvtlsU 215573 Normal
## 3 1YMBg7rOjxzbya0fPOYfNX 198286 Normal
## 4 4skbQNtyjy8A7mo8oqe2oD 192455 Normal
## 5 3Yvi5NkUrSppVwrMHYkB6u 97667 Normal
## 6 00cBcYOlnHoXX9ver3cmdE 225680 NormalFinding Possible and Missing Combinations.
possible_combinations2 <- expand.grid(genre = unique(data$genre),
duration_ms = unique(data$duration_ms))
print(head(possible_combinations))## artistName msPlayed
## 1 Project AER 119999
## 2 AURORA 119999
## 3 DJ Snake 119999
## 4 Chris Lake 119999
## 5 James Horner 119999
## 6 JP Saxe 119999missing_combinations2 <- anti_join(possible_combinations2, data, by = c("genre" , "duration_ms"))
print(head(missing_combinations2))## genre duration_ms
## 1 art pop 120000
## 2 edm 120000
## 3 bass house 120000
## 4 orchestral soundtrack 120000
## 5 alt z 120000
## 6 alternative metal 120000data_for_combination2 <-head(data,8000)
combination_counts2 <- data_for_combination2 %>%
group_by(genre, duration_ms) %>%
summarise(count = n(), .groups = 'drop')
print(combination_counts2)## # A tibble: 4,252 × 3
## genre duration_ms count
## <chr> <int> <int>
## 1 a cappella 269636 2
## 2 abstract 465200 2
## 3 abstract beats 73898 2
## 4 abstract hip hop 283747 1
## 5 acoustic opm 244800 2
## 6 acoustic pop 186584 2
## 7 acoustic pop 189613 2
## 8 acoustic pop 198853 2
## 9 acoustic pop 226680 2
## 10 acoustic pop 231587 1
## # ℹ 4,242 more rowsAfter calculating the frequencies we got different counts for different genre so we can analyze which genre have how many combinations with duration_ms.
Group by 3
data3<-head(data,40)data_group3 <- data %>%
group_by(genre) %>%
summarise(mean_loudness= mean(loudness),
)
print(data_group3)## # A tibble: 523 × 2
## genre mean_loudness
## <chr> <dbl>
## 1 a cappella -37.8
## 2 abstract -11.9
## 3 abstract beats -8.69
## 4 abstract hip hop -6.46
## 5 acoustic opm -8.13
## 6 acoustic pop -9.25
## 7 adult standards -12.4
## 8 aesthetic rap -9.71
## 9 afghan pop -10.6
## 10 afrobeats -7.23
## # ℹ 513 more rowssummary(data_group3)## genre mean_loudness
## Length:523 Min. :-37.841
## Class :character 1st Qu.: -9.626
## Mode :character Median : -7.528
## Mean : -8.528
## 3rd Qu.: -5.728
## Max. : -0.503ggplot(data_group3, aes(genre,mean_loudness)) + geom_col(size=3.5, color="#3b5998") + labs(title = "Mean loudness played according to genre", 'genre','mean_loudness')+theme(axis.text.x = element_text(angle = 90))## Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
## ℹ Please use `linewidth` instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.
data_group3 <- data_group3 %>%
mutate(exp_prob3 = (mean_loudness / sum(mean_loudness)))
print(data_group3)## # A tibble: 523 × 3
## genre mean_loudness exp_prob3
## <chr> <dbl> <dbl>
## 1 a cappella -37.8 0.00848
## 2 abstract -11.9 0.00267
## 3 abstract beats -8.69 0.00195
## 4 abstract hip hop -6.46 0.00145
## 5 acoustic opm -8.13 0.00182
## 6 acoustic pop -9.25 0.00207
## 7 adult standards -12.4 0.00279
## 8 aesthetic rap -9.71 0.00218
## 9 afghan pop -10.6 0.00238
## 10 afrobeats -7.23 0.00162
## # ℹ 513 more rowsanomaly3 <- data_group3 %>%
filter(exp_prob3 == min(exp_prob3)) %>%
pull(genre)
print(anomaly3)## [1] "j-idol"data3 <- data %>%
mutate(anomaly_3 = ifelse(genre == anomaly3, "Anomaly", "Normal"))
print(head(data3$anomaly_3 == "Anomaly") )## [1] FALSE FALSE FALSE FALSE FALSE FALSEIn this group genre is considered as anomaly as it is having less probability.
Finding Possible and Missing Combinations.
possible_combinations3 <- expand.grid(genre = unique(data$genre),
loudness = unique(data$loudness))
print(head(possible_combinations3))## genre loudness
## 1 ambient guitar -17.984
## 2 art pop -17.984
## 3 edm -17.984
## 4 bass house -17.984
## 5 orchestral soundtrack -17.984
## 6 alt z -17.984missing_combinations3 <- anti_join(possible_combinations3, data, by = c("genre" , "loudness"))
print(head(missing_combinations3))## genre loudness
## 1 art pop -17.984
## 2 edm -17.984
## 3 bass house -17.984
## 4 orchestral soundtrack -17.984
## 5 alt z -17.984
## 6 alternative metal -17.984data_for_combination3 <-head(data,8000)
combination_counts3 <- data_for_combination3 %>%
group_by(genre, loudness) %>%
summarise(count = n(), .groups = 'drop')
print(head(combination_counts3))## # A tibble: 6 × 3
## genre loudness count
## <chr> <dbl> <int>
## 1 a cappella -37.8 2
## 2 abstract -11.9 2
## 3 abstract beats -8.69 2
## 4 abstract hip hop -6.46 1
## 5 acoustic opm -8.13 2
## 6 acoustic pop -12.6 2After calculating the frequencies we got different counts for different genre so we can analyze which genre have how many combinations with loudness.
Conclusion
As I have categorical variables with a large number of unique categories, creating visualizations became cluttered and challenging to interpret. So it is more appropriate to aggregate or group the categories before visualization.
Due to this reason I am unable to plot the graphs with showing all the categorical names.
There is no clear pattern or relationship between the categorical and numerical variables. In these instances, creating a visualization may not reveal any meaningful insights.