---
tags: AT_DH_2022
---

# Jagatud märkmed - ATDH-2022-2 (HVEE.00.016)

Püsilink: https://hackmd.io/@OGZFb2mRSA65ybyn4hSfkQ/BJ1DbApgs/edit


Pildi jagamiseks paremklõps pildi peal (all paremal nurgas) -> copy image.

Ja CTRL+V (CMD+V) siin tekstikastis. Hackmd laeb faili üles ja siis kuvab lingi markdown formaadis. Joonise järgi saab kirjutada oma nime.

Joonis 1. Näitepilt

## 13.09.2022

# Kui palju on jõululaule?
# Kas jõululood on stabiilsemad (muutuvad vähem)?
# Kas jõulud pikenevad?

# Tegelikult peaks teadma, kust andmed pärinevad.


# Kui palju on ühehitibände - kas viimasel ajal rohkem?
#1975-1985 hüpe - oli detsentraliseeritum? ühe-hitibändid?

billboard_data %>%
  mutate(primary=str_remove(str_extract(artist,"^(.*)( featuring)")," featuring")) %>% 
  group_by(primary) %>% 
  summarise(n=n())

`# For example first and last year for each artist
billboard_data %>%
group_by(artist) %>%
summarise(n=n(),firstyear=min(year),lastyear=max(year)) %>%
arrange(desc(n)) %>%
filter(lastyear>2010)
`

billboard_data %>% 
  filter(artist=="madonna"|artist=="elton john"|artist=="mariah carey"|artist=="stevie wonder") %>% 
  ggplot(aes(x=year,y=-rank))+
  geom_point()+
  #stat_smooth(method="loess")+
  facet_wrap(~artist)


billboard_data %>% 
  filter(str_detect(lyrics,"baby|yeah|love|war|peace|kill|darling|phone|car")) %>% 
  mutate(found=str_extract(lyrics,"baby|yeah|love|war|peace|kill|darling|phone|car")) %>% 
  ggplot(aes(x=year,y=-rank))+
  stat_smooth(method="lm")+
  geom_point(size=0.1)+
  facet_wrap(~found)+
  theme_classic()

billboard_data %>% 
  filter(str_detect(lyrics,"love",negate=T)) %>% 
  #mutate(found=str_extract(lyrics,"baby|yeah|love|war|peace|kill|darling|phone|car")) %>% 
  ggplot(aes(x=year,y=-rank))+
  stat_smooth(method="lm")+
  geom_point(size=0.1)+
  #facet_wrap(~found)+
  theme_classic()

billboard_data %>% 
  filter(str_detect(lyrics,"wind|earth|fire|water")) %>% 
  mutate(found=str_extract(lyrics,"wind|earth|fire|water")) %>% 
  ggplot(aes(x=year,y=-rank))+
  stat_smooth(method="lm")+
  geom_point(size=0.1)+
  facet_wrap(~found)+
  theme_classic()

billboard_data %>% 
  filter(str_detect(lyrics,"christmas")) %>% 
  mutate(found=str_extract(lyrics,"christmas")) %>% 
  ggplot(aes(x=year,y=-rank))+
  stat_smooth(method="lm")+
  geom_point(size=0.1)+
  facet_wrap(~found)+
  theme_classic()

billboard_data %>% 
  group_by(artist) %>% 
  summarise(n=n(),firstyear=min(year),lastyear=max(year)) %>% 
  arrange(desc(n)) %>% 
  head(50) %>% 
  arrange(desc(firstyear)) %>%  # here we sort them by the first year before we set the factor levels
  mutate(artist=factor(artist,levels=unique(artist))) %>% 
  ggplot(aes(y=artist))+
  geom_segment(aes(yend=artist,x=firstyear,xend=lastyear),size=3)+
  geom_text(aes(label=n,x=1963))+
  theme_minimal()

## 16.09.2022

Kui tahame sõnu võrrelda, siis asendame rank telje word teljega ja siis on samad sõnad kõrvuti. Sõnad võib põhimõtteliselt järjestada rank järgi (fct_reorder funktsioon) - aga kuna igal arte ei ole nii päris puhas.

# Top words for 4 selected artists
artists_topwords %>% 
  filter(n_rank<26) %>% 
  filter(artist=="madonna"|artist=="rihanna"|artist=="aretha franklin"|artist=="britney spears") %>% 
  #filter(word=="love"|word=="you"|word=="baby"|word=="time") %>% 
  ggplot(aes(x=word,y=artist,fill=word,label=word))+
  geom_tile(alpha=0.4)+
  geom_text()+
  guides(fill=FALSE)+
  coord_flip()
 

# Top words for 4 selected artists
artists_topwords %>% 
  filter(n_rank<26) %>% 
  filter(artist=="madonna"|artist=="rihanna"|artist=="aretha franklin"|artist=="britney spears") %>% 
  #filter(word=="love"|word=="you"|word=="baby"|word=="time") %>% 
  ggplot(aes(x=fct_reorder(word,n_rank),y=artist,fill=word,label=word))+
  geom_tile(alpha=0.4)+
  geom_text()+
  guides(fill=FALSE)+
  coord_flip()





Siin on lood järjestatud viimase asukoha järgi, teeb puhtama graafiku.

billboard_tokens %>% 
  filter(word=="party") %>% 
  ggplot(aes(x=loc_perc,y=fct_reorder(song,loc_perc,.fun=max)))+
  geom_point()

"Begin"

"Coming"