Quotation analysis

top_quotes_categorized_local <- read_sheet(ss="1Lx2nhdwdgJsxrn047Wnre1S3GsrQAQ1OFZ3UOwTVQfE",sheet="Top quoted Categorized") %>% 
  rename(group=`group 1`,subgroup=`group 1.1`) %>%
  mutate(group=if_else(!is.na(status) & status=="Pois","Pois",group),subgroup=if_else(!is.na(status) & status=="Pois","Pois",subgroup)) %>%
  mutate(subgroup=coalesce(subgroup,group)) %>%
  filter(!is.na(group))  

The googlesheets4 package is requesting access to your Google account.
Select a pre-authorised account or enter '0' to obtain a new token.
Press Esc/Ctrl + C to cancel.

1: jiemakel@gmail.com

1

Auto-refreshing stale OAuth token.
✔ Reading from Lainaukset ja valta -aineisto.
✔ Range ''Top quoted Categorized''.

random_quotes_categorized_local <- read_sheet(ss="1Lx2nhdwdgJsxrn047Wnre1S3GsrQAQ1OFZ3UOwTVQfE",sheet="random quotes") %>% 
  union_all(read_sheet(ss="1Lx2nhdwdgJsxrn047Wnre1S3GsrQAQ1OFZ3UOwTVQfE",sheet="random quotes 2")) %>%
  rename(group=`group 1`,subgroup=`group 1.1`) %>%
  mutate(group=if_else(!is.na(status) & status=="Pois","Pois",group),subgroup=if_else(!is.na(status) & status=="Pois","Pois",subgroup)) %>%
  mutate(subgroup=coalesce(subgroup,group)) %>%
  filter(!is.na(group))

✔ Reading from Lainaukset ja valta -aineisto.
✔ Range ''random quotes''.
✔ Reading from Lainaukset ja valta -aineisto.
✔ Range ''random quotes 2''.

top_quotes_categorized_a <- top_quotes_categorized_local %>%
  copy_to_a(con, unique_indexes=list(c("canonical_name","after","before")))

top_quotes_categorized_c <- top_quotes_categorized_a %>%
  compute_c()

random_quotes_categorized_a <- random_quotes_categorized_local %>%
  copy_to_a(con)

random_quotes_categorized_c <- random_quotes_categorized_a %>%
  compute_c()

yearly_quotes_local <- top_quotes_categorized_c %>%
  select(-c(indirect:canonical_joined_role_all)) %>%
  inner_join(quote_author_names_to_canonical_names_c) %>%
  inner_join(quote_authors_c) %>%
  inner_join(q_qa_c) %>%
  right_join(
    article_types_c %>% 
      filter(type=="Domestic general/political/economic news") %>% 
      inner_join(quotes_c)) %>%
  inner_join(articles_c) %>%
  filter(is.na(after) | year_created>=after, is.na(before) | year_created<before) %>%
  count(canonical_name,group,subgroup,direct,media,year_created) %>%
  ungroup() %>%
  collect() %>%
  mutate(direct=if_else(direct==1,"direct","indirect"))

Joining, by = "canonical_name"Joining, by = "name"Joining, by = "qa_id"Joining, by = "a_id"Joining, by = "q_id"Joining, by = "a_id"

Absolute view of quotes in domestic general/political/economic news

yearly_quotes_local %>%
  group_by(media,group,direct,year_created) %>%
  summarize(n=sum(n),.groups="drop") %>%
  ggplot(aes(x=year_created,y=n,fill=group)) +
  geom_col(width=1) + 
  facet_grid(media~direct) + 
  theme_hsci_discrete(base_family="Arial") +
  theme(legend.position = 'bottom')

Proportional view of quotes in domestic general/political/economic news

yearly_quotes_local %>%
  group_by(media,group,direct,year_created) %>%
  summarize(n=sum(n),.groups="drop") %>%
  ggplot(aes(x=year_created,y=n,fill=group)) +
  geom_col(position='fill',width=1) + 
  facet_grid(media~direct) + 
  theme_hsci_discrete(base_family="Arial") +
  theme(legend.position = 'bottom')

Proportions of quotes categorized in the different main categories in domestic general/political/economic news

yearly_quotes_local %>%
  group_by(media,group,direct,year_created) %>%
  summarize(n=sum(n),.groups="drop") %>%
  group_by(media,direct,year_created) %>%
  mutate(prop=n/sum(n)) %>%
  ungroup() %>%
  filter(!is.na(group)) %>%
  ggplot(aes(x=year_created,y=prop,color=group)) +
  geom_step() + 
  facet_grid(media~direct) + 
  theme_hsci_discrete(base_family="Arial") +
  scale_y_continuous(labels=scales::percent_format(accuracy=1)) +
  theme(legend.position = 'bottom')

Proportions of quotes categorized in the different subcategories in domestic general/political/economic news

yearly_quotes_local %>%
  group_by(subgroup,media,direct,year_created) %>%
  summarize(n=sum(n),.groups="drop") %>%
  group_by(media,direct,year_created) %>%
  mutate(prop=n/sum(n)) %>%
  ungroup() %>%
  filter(!is.na(subgroup)) %>%
  ggplot(aes(x=year_created,y=prop,color=subgroup)) +
  geom_step() + 
  facet_grid(media~direct) + 
  theme_hsci_discrete(base_family="Arial") +
  scale_y_continuous(labels=scales::percent_format(accuracy=1)) +
  theme(legend.position = 'bottom')

Proportions of quotes for top sources in domestic general/political/economic news

yearly_quotes_local %>%
  group_by(canonical_name,group,media,direct,year_created) %>%
  summarize(n=sum(n),.groups="drop") %>%
  group_by(media,direct,year_created) %>%
  mutate(prop=n/sum(n)) %>%
  ungroup() %>%
  filter(group!="Pois") %>%
  mutate(canonical_name=fct_lump_n(canonical_name,12,w=n)) %>%
  filter(canonical_name!="Other") %>%
  ggplot(aes(x=year_created,y=prop,color=canonical_name)) +
  geom_step() + 
  facet_grid(media~direct) + 
  theme_hsci_discrete(base_family="Arial") +
  scale_y_continuous(labels=scales::percent_format(accuracy=1)) +
  theme(legend.position = 'bottom')

Overall proportions

d <- random_quotes_categorized_local %>%
  mutate(period=year(time_created)>=2009, direct=if_else(direct==0,"indirect","direct")) %>%
  count(group,media,direct,period) %>%
  group_by(media,direct,period) %>%
  group_modify(~bind_cols(.x,scimp_bmde(.x$n, p=1))) %>%
  mutate(prop=n/sum(n),rest=TRUE) %>%
  ungroup() %>%
  right_join(yearly_quotes_local %>%
    mutate(period=year_created>=2009) %>%
    group_by(group,media,direct,period) %>%
    summarize(n=sum(n),.groups="drop") %>%
    group_by(media,direct,period) %>%
    mutate(prop=n/sum(n)) %>%
    ungroup() %>%
    mutate(rest=is.na(group)),by=c("media","direct","period","rest")) %>%
  mutate(group=coalesce(group.x,group.y), prop=if_else(is.na(group.y),prop.y*prop.x,prop.y),upper_limit=if_else(is.na(group.y),prop.y*upper_limit,prop.y), lower_limit=if_else(is.na(group.y),prop.y*lower_limit,prop.y)) %>%
  group_by(media,direct,period,group) %>% 
  summarize(prop=sum(prop),upper_limit=sum(upper_limit),lower_limit=sum(lower_limit),.groups="drop") %>%
  filter(group!="Pois") %>%
  group_by(media,direct,period) %>%
  mutate(prop=prop/sum(prop), upper_limit=upper_limit/sum(upper_limit), lower_limit=lower_limit/sum(lower_limit)) %>%
  ungroup() %>%
  mutate(period=if_else(period==FALSE,"1999-2008","2009-2018"))

d2 <- random_quotes_categorized_local %>%
  mutate(period=year(time_created)>=2009, direct=if_else(direct==0,"indirect","direct")) %>%
  count(subgroup,media,direct,period) %>%
  group_by(media,direct,period) %>%
  group_modify(~bind_cols(.x,scimp_bmde(.x$n, p=1))) %>%
  mutate(prop=n/sum(n),rest=TRUE) %>%
  ungroup() %>%
  right_join(yearly_quotes_local %>%
    mutate(period=year_created>=2009) %>%
    group_by(subgroup,media,direct,period) %>%
    summarize(n=sum(n),.groups="drop") %>%
    group_by(media,direct,period) %>%
    mutate(prop=n/sum(n)) %>%
    ungroup() %>%
    mutate(rest=is.na(subgroup)),by=c("media","direct","period","rest")) %>%
  mutate(subgroup=coalesce(subgroup.x,subgroup.y), prop=if_else(is.na(subgroup.y),prop.y*prop.x,prop.y),upper_limit=if_else(is.na(subgroup.y),prop.y*upper_limit,prop.y), lower_limit=if_else(is.na(subgroup.y),prop.y*lower_limit,prop.y)) %>%
  group_by(media,direct,period,subgroup) %>% 
  summarize(prop=sum(prop),upper_limit=sum(upper_limit),lower_limit=sum(lower_limit),.groups="drop") %>%
  filter(subgroup!="Pois") %>%
  group_by(media,direct,period) %>%
  mutate(prop=prop/sum(prop), upper_limit=upper_limit/sum(upper_limit), lower_limit=lower_limit/sum(lower_limit)) %>%
  ungroup() %>%
  mutate(period=if_else(period==FALSE,"1999-2008","2009-2018"))

By group

d %>% 
  ggplot(aes(x=group,color=period)) +
  geom_point(aes(y=prop)) + 
  geom_errorbar(aes(ymin=lower_limit,ymax=upper_limit)) +
  facet_grid(media~direct) +
  coord_flip() +
  theme_hsci_discrete(base_family="Arial") +
  scale_y_continuous(labels=scales::percent_format(accuracy=1)) +
  theme(legend.position="bottom") +
  xlab("Group") +
  ylab("Proportion")

d %>%
  ggplot(aes(x=group,group=media,color=media)) +
  geom_point(aes(y=prop),position=position_dodge(width=1)) + 
  geom_errorbar(aes(ymin=lower_limit,ymax=upper_limit),position=position_dodge(width=1)) +
  facet_grid(period~direct) +
  coord_flip() +
  theme_hsci_discrete(base_family="Arial") +
  scale_y_continuous(labels=scales::percent_format(accuracy=1)) +
  theme(legend.position="bottom") +
  xlab("Group") +
  ylab("Proportion")

d %>%
  ggplot(aes(x=group,group=media,color=direct)) +
  geom_point(aes(y=prop),position=position_dodge(width=1)) + 
  geom_errorbar(aes(ymin=lower_limit,ymax=upper_limit),position=position_dodge(width=1)) +
  facet_grid(media~period) +
  coord_flip() +
  theme_hsci_discrete(base_family="Arial") +
  scale_y_continuous(labels=scales::percent_format(accuracy=1)) +
  theme(legend.position="bottom") +
  xlab("Subgroup") +
  ylab("Proportion")

By subgroup

d2 %>% 
  ggplot(aes(x=subgroup,color=period)) +
  geom_point(aes(y=prop)) + 
  geom_errorbar(aes(ymin=lower_limit,ymax=upper_limit)) +
  facet_grid(media~direct) +
  coord_flip() +
  theme_hsci_discrete(base_family="Arial") +
  scale_y_continuous(labels=scales::percent_format(accuracy=1)) +
  theme(legend.position="bottom") +
  xlab("Subgroup") +
  ylab("Proportion")

d2 %>%
  ggplot(aes(x=subgroup,group=media,color=media)) +
  geom_point(aes(y=prop),position=position_dodge(width=1)) + 
  geom_errorbar(aes(ymin=lower_limit,ymax=upper_limit),position=position_dodge(width=1)) +
  facet_grid(period~direct) +
  coord_flip() +
  theme_hsci_discrete(base_family="Arial") +
  scale_y_continuous(labels=scales::percent_format(accuracy=1)) +
  theme(legend.position="bottom") +
  xlab("Subgroup") +
  ylab("Proportion")

d2 %>%
  ggplot(aes(x=subgroup,group=media,color=direct)) +
  geom_point(aes(y=prop),position=position_dodge(width=1)) + 
  geom_errorbar(aes(ymin=lower_limit,ymax=upper_limit),position=position_dodge(width=1)) +
  facet_grid(media~period) +
  coord_flip() +
  theme_hsci_discrete(base_family="Arial") +
  scale_y_continuous(labels=scales::percent_format(accuracy=1)) +
  theme(legend.position="bottom") +
  xlab("Subgroup") +
  ylab("Proportion")