Sources of the Kalevala

works_of_interest_l <- tribble(
  ~book, ~work, ~pattern, ~year, ~n_verses_prev, ~publication_order,
  "Väitöskirjan aineksia (1827)", "Väitöskirjan aineksia (1827)", "^kr00031", 1827, 95, 1,
  "Kantele I (1829)", "Kantele I (1829)", "^kr00038", 1829, NA, 2,
  "Kantele II (1829)", "Kantele II (1829)", "^kr00039", 1829, NA, 3,
  "Kantele III (1830)", "Kantele III (1830)", "^kr00040", 1830, NA, 4,
  "Kantele IV (1831)", "Kantele IV (1831)", "^kr00041", 1831, NA, 5,
  "Sikermä-Kalevala (1833)", "Sikermä-Kalevala (1833) - Lemminkäinen", "^kr00032", 1833, 825, 6,
  "Sikermä-Kalevala (1833)", "Sikermä-Kalevala (1833) - Väinämöinen", "^kr00033", 1833, 1867, 6,
  "Sikermä-Kalevala (1833)", "Sikermä-Kalevala (1833) - Naimakansan virsiä Vuokkiniemeltä", "^kr00034", 1833, 499, 6,
  "Runokokous Väinämöisestä (1833)","Runokokous Väinämöisestä (1833)", "^kr00035", 1833, 5052, 7,
  "Lisiä runokokoukseen Väinämöisestä (1834)", "Lisiä runokokoukseen Väinämöisestä (1834)", "^kr00036", 1834, NA, 8,
  "Vanha Kalevala (1835)", "Vanha Kalevala (1835)", "^vkalevala", 1835, 12078, 9,
  "Lisiä Vanhaan Kalevalaan (1836-1837)", "Lisiä Vanhaan Kalevalaan (1836-1837)", "^kr00037", 1836, NA, 10,
  "Kanteletar (1840)", "Kanteletar (1840)", "^kt", 1840, NA, 11,
  "Kalevala (1849)", "Kalevala (1849) - Luominen ja Väinämöisen syntymä", "^kalevala0[1-2]", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Aino-runo", "^kalevala0[3-5]", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Väinämöisen ja Ilmarisen matkat Pohjolaan", "^kalevala(0[6-9]|10)", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Lemminkäinen I", "^kalevala1[1-5]", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Väinämöisen veneenveisto ja Pohjolasta kosinta I", "^kalevala1[6-9]", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Häärunot", "^kalevala2[0-5]", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Lemminkäinen II", "^kalevala(2[6-9]|30)", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Kullervo", "^kalevala3[1-6]", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Kultaneito ja Pohjolasta kosinta II", "^kalevala3[7-8]", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Matkalla sammon ryöstöön", "^kalevala(39|4[0-1])", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Taistelu sammosta", "^kalevala4[2-3]", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Kanteleen soitto", "^kalevala44", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Kalevalan ja Pohjolan välinen taistelu", "^kalevala4[5-9]", 1849, 22795, 12,
  "Kalevala (1849)", "Kalevala (1849) - Väinämöisen lähtö", "^kalevala50", 1849, 22795, 12,
) %>% mutate(display_order=row_number())

works_of_interest <- works_of_interest_l %>%
  copy_to_a(con, temporary=FALSE, overwrite=TRUE, name="works_of_interest", unique_indexes=list(c("work")))

work_parts_of_interest <- works_of_interest_l %>% 
  mutate(p_id=map(pattern,~poems %>% 
                    filter(str_detect(nro,.x)) %>% 
                    pull(p_id))) %>% 
  unnest(p_id) %>% 
  select(-pattern) %>%
  copy_to_a(con, temporary=FALSE, overwrite=TRUE, name="work_parts_of_interest", indexes=list(c("work")), unique_indexes=list(c("p_id")))

work_verses <- work_parts_of_interest %>% 
  select(work,p_id) %>%
  inner_join(verse_poem %>% 
               select(p_id, pos, v_id)) %>% 
  inner_join(verses %>% 
               filter(type=="V") %>% 
               select(v_id)) %>%
  compute_a(temporary=FALSE, overwrite=TRUE, name="work_verses", indexes=list(c("v_id"),c("work")), unique_indexes=list(c("p_id","pos","v_id")))

poems_of_interest <- poems %>% 
  filter(collection %in% c("skvr","jr")) %>%
  inner_join(p_year %>% filter(year>=1700,year<1849)) %>%
  compute_a(temporary=FALSE, overwrite=TRUE, name="poems_of_interest", unique_indexes=list(c("p_id")))

poem_verses <- poems_of_interest %>%
  select(p_id) %>%
  inner_join(verse_poem %>% 
               select(p_id, pos, v_id)) %>% 
  inner_join(verses %>% 
               filter(type=="V") %>% 
               select(v_id)) %>%
  compute_a(temporary=FALSE, overwrite=TRUE, name="poems_verses", indexes=list(c("v_id")), unique_indexes=list(c("p_id","pos","v_id")))

poem_verse_clusts <- poem_verses %>% 
  inner_join(v_clust %>% inner_join(v_clusterings %>% filter(name=="loose"))) %>% 
  distinct(p_id,clust_id) %>%
  compute_a(temporary=FALSE, overwrite=TRUE, name="poem_verse_clusts", indexes=list(c("clust_id")), unique_indexes=list(c("p_id","clust_id")))

work_poem_sims <- work_verses %>% 
  rename(v1_id=v_id, p1_id=p_id, pos1=pos) %>%
  inner_join(v_sim) %>%
  inner_join(poem_verses %>% 
               rename(v2_id=v_id, p2_id=p_id, pos2=pos)) %>% 
  compute_a(temporary=FALSE,overwrite=TRUE, name="work_poem_sims", indexes=list(c("work"),c("p1_id","pos1"),c("p2_id","pos2")))

work_work_sims <- work_verses %>% 
  rename(v1_id=v_id, p1_id=p_id, pos1=pos, work1=work) %>%
  inner_join(v_sim) %>% 
  inner_join(work_verses %>% rename(v2_id=v_id, p2_id=p_id, pos2=pos, work2=work)) %>% 
  filter(work1!=work2) %>%
  compute_a(temporary=FALSE,overwrite=TRUE, name="work_work_sims", indexes=list(c("sim_cos"),c("work1"),c("work2"),c("p1_id","pos1"),c("p2_id","pos2")))

work_verse_clusts <- work_verses %>% 
  inner_join(v_clust %>% inner_join(v_clusterings %>% filter(name=="loose"))) %>% 
  distinct(work,clust_id) %>%
  compute_a(temporary=FALSE, overwrite=TRUE, name="work_verse_clusts", indexes=list(c("clust_id")), unique_indexes=list(c("work","clust_id")))

work_verse_clust_origins <- work_verse_clusts %>%
  inner_join(works_of_interest) %>%
  group_by(clust_id) %>%
  filter(publication_order==min(publication_order)) %>%
  ungroup() %>%
  select(origin_work=work, clust_id) %>%
  compute_a(temporary=FALSE, overwrite=TRUE, name="work_verse_clust_origins", indexes=list(c("origin_work"), c("clust_id")))

poems_to_cols <- poems_of_interest %>%
  inner_join(p_col) %>%
  inner_join(collectors %>% rename(col_name=name)) %>%
  inner_join(p_pl) %>%
  inner_join(places %>% select(pl_id, par_id)) %>%
  inner_join(places %>% filter(type=="county") %>% rename(pl_name=name), join_by(par_id==pl_id)) %>%
  mutate(pl_name=case_match(pl_name,
                            c("Uusimaa","Varsinais-Suomi","Etelä-Pohjanmaa") ~ "Uusimaa/Varsinais-Suomi/Etelä-Pohjanmaa",
                            c("Keski-Suomi","Pohjois-Savo","Etelä-Savo") ~ "Keski-Suomi/Savo",
                            c("Satakunta","Häme") ~ "Satakunta/Häme",
                            c("Peräpohjola","Pohjois-Pohjanmaa","Kainuu") ~ "Peräpohjola/Pohjois-Pohjanmaa/Kainuu",
                            c("Etelä-Karjala") ~ "Etelä-Karjala",
                            c("Aunus") ~ "Aunus",
                            c("Länsi-Inkeri","Keski-Inkeri","Itä- ja Pohjois-Inkeri") ~ "Inkeri",
                            c("Laatokan Karjala (Raja-Karjala)") ~ "Laatokan Karjala",
                            c("Viena") ~ "Viena",
                            c("Pohjois-Karjala") ~ "Pohjois-Karjala")) %>%
  distinct(p_id, year, col_name, pl_name) %>%
  compute_a(temporary=FALSE, overwrite=TRUE, name="poems_to_cols",indexes=list(c("p_id"), c("year")))

work_parts_of_interest %>% 
  inner_join(poem_stats) %>% 
  group_by(publication_order,work, n_verses_prev) %>% 
  summarise(poems=n(), verses=sum(nverses), all_lines=sum(n_all_lines)) %>% 
  arrange(publication_order,work) %>%
  relocate(n_verses_prev, .before=verses)

Works under analysis

work_parts_of_interest %>% 
  inner_join(poem_stats) %>% 
  group_by(book) %>% 
  summarise(poems=n(), verses=sum(nverses), all_lines=sum(n_all_lines),.groups="drop") %>% 
  inner_join(work_verse_clusts %>% inner_join(works_of_interest) %>% count(book, name="verse_types")) %>%
  left_join(works_of_interest %>% group_by(book) %>% summarise(display_order=min(display_order))) %>%
  arrange(display_order) %>%
  select(book,poems,verses,verse_types) %>%
  gt(rowname_col = "book") %>%
  fmt_integer(c(poems, verses, verse_types), sep_mark = " ") %>%
  cols_label(verses="Verses",poems="Poems",verse_types="Verse types")

	Poems	Verses	Verse types
Väitöskirjan aineksia (1827)	1	98	80
Kantele I (1829)	25	1 490	1 207
Kantele II (1829)	54	1 814	1 499
Kantele III (1830)	16	1 493	1 223
Kantele IV (1831)	27	1 609	1 384
Sikermä-Kalevala (1833)	7	4 754	3 113
Runokokous Väinämöisestä (1833)	17	8 006	4 681
Lisiä runokokoukseen Väinämöisestä (1834)	1	4 040	2 700
Vanha Kalevala (1835)	64	15 249	8 121
Lisiä Vanhaan Kalevalaan (1836-1837)	34	21 399	12 319
Kanteletar (1840)	676	23 133	14 071
Kalevala (1849)	50	22 796	16 788

Work development

work_nodes_l <- works_of_interest_l %>% 
  mutate(id=row_number() - 1)

work_verse_network_l <- work_verse_clusts %>% 
  inner_join(works_of_interest %>% select(work, publication_order)) %>%
  right_join(work_verse_clusts %>% 
    inner_join(works_of_interest %>% select(work, publication_order)), join_by(clust_id, publication_order < publication_order)) %>%
  mutate(work.x=coalesce(work.x,work.y), publication_order.x=coalesce(publication_order.x,publication_order.y)) %>%
  group_by(clust_id, work.y) %>%
  filter(publication_order.x==max(publication_order.x)) %>%
  ungroup() %>%
  inner_join(work_verse_clust_origins) %>%
  count(origin_work, work.x,work.y) %>%
  collect() %>%
  inner_join(work_nodes_l %>% select(work.x=work,id.x=id)) %>%
  inner_join(work_nodes_l %>% select(work.y=work,id.y=id)) %>% 
  mutate(id.x=if_else(id.y==id.x,14+id.x,id.x))

work_nodes_l <- work_nodes_l %>%
  union_all(work_nodes_l %>% mutate(id=id+14))

library(networkD3)
sankeyNetwork(Nodes=work_nodes_l,Links=work_verse_network_l, NodeID='work', Source='id.x', Target='id.y',Value='n', fontSize = 12, width=2048, height=2048, nodeWidth = 30, LinkGroup = 'origin_work', units="verses", colourScale=JS("d3.scaleOrdinal(d3.schemeCategory20c);"))

work_nodes_l <- works_of_interest_l %>% 
  mutate(id=row_number() - 1)

n_works <- work_nodes_l %>% count() %>% pull()

work_verse_network_l <- work_verse_clusts %>% 
  inner_join(works_of_interest %>% select(work, publication_order)) %>%
  right_join(work_verse_clusts %>% 
    inner_join(works_of_interest %>% select(work, publication_order)), join_by(clust_id, publication_order < publication_order)) %>%
  mutate(work.x=coalesce(work.x,work.y), publication_order.x=coalesce(publication_order.x,publication_order.y)) %>%
  group_by(clust_id, work.y) %>%
  filter(publication_order.x==max(publication_order.x)) %>%
  ungroup() %>%
  inner_join(work_verse_clust_origins) %>%
  count(work.x,work.y) %>%
  collect() %>%
  inner_join(work_nodes_l %>% select(work.x=work,id.x=id)) %>%
  inner_join(work_nodes_l %>% select(work.y=work,id.y=id)) %>% 
  mutate(id.x=if_else(id.y==id.x,n_works+id.x,id.x))

Joining with `by = join_by(work)`Joining with `by = join_by(work)`Joining with `by = join_by(clust_id)`Joining with `by = join_by(work.x)`Joining with `by = join_by(work.y)`

work_nodes_l <- work_nodes_l %>%
  union_all(work_nodes_l %>% mutate(id=id+n_works))

library(networkD3)
sankeyNetwork(Nodes=work_nodes_l,Links=work_verse_network_l, NodeID='work', Source='id.x', Target='id.y',Value='n', fontSize = 12, width=2048, height=2048, nodeWidth = 30, LinkGroup = 'work.x', units="verses", colourScale=JS("d3.scaleOrdinal(d3.schemeCategory20c);"))

Links is a tbl_df. Converting to a plain data frame.
Nodes is a tbl_df. Converting to a plain data frame.

work_nodes_l <- works_of_interest_l %>% 
  mutate(id=row_number() - 1)

Error: object 'works_of_interest_l' not found

library(networkD3)
sankeyNetwork(Nodes=work_nodes_l,Links=work_verse_network_l, NodeID='work', Source='id.x', Target='id.y',Value='n', fontSize = 12, width=2048, height=2048, nodeWidth = 30, LinkGroup = 'work.x', units="verses", colourScale=JS("d3.scaleOrdinal(d3.schemeCategory20c);"))

work_verse_clusts %>% 
  inner_join(work_verse_clust_origins) %>%
  inner_join(works_of_interest %>% select(origin_work=work,publication_order)) %>%
  inner_join(works_of_interest %>% select(work,publication_order), join_by(work)) %>%
  filter(publication_order.x<publication_order.y | work == origin_work) %>%
  count(origin_work, work, publication_order.x, publication_order.y) %>%
  group_by(work) %>%
  mutate(prop=n/sum(n)) %>%
  ungroup() %>%
  collect() %>%
  mutate(origin_work=fct_reorder(origin_work, publication_order.x),work=fct_reorder(work, publication_order.y)) %>%
  ggplot(aes(x=origin_work,y=work,fill=prop)) +
  geom_tile() +
  theme_hsci_continuous() +
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

Joining with `by = join_by(clust_id)`Joining with `by = join_by(origin_work)`

Origins of verse types in parts of the Kalevala in terms of earlier works

work_verse_clusts %>% 
  inner_join(work_verse_clust_origins) %>%
  inner_join(works_of_interest %>% select(origin_work=work,publication_order)) %>%
  inner_join(works_of_interest %>% select(work,publication_order), join_by(work)) %>%
  filter(publication_order.x<publication_order.y | work == origin_work) %>%
  count(origin_work, work, publication_order.x, publication_order.y) %>%
  group_by(work) %>%
  mutate(prop=n/sum(n)) %>%
  ungroup() %>%
  filter(publication_order.y==12,origin_work!=work) %>%
  collect() %>%
  inner_join(works_of_interest_l %>% mutate(order=row_number()) %>% select(work,order)) %>%
  mutate(origin_work=fct_reorder(origin_work, publication_order.x),work=fct_reorder(work, order,.desc=TRUE)) %>%
  ggplot(aes(x=work,y=prop,fill=origin_work)) +
  geom_col() +
  scale_y_continuous(labels=scales::percent) +
  coord_flip() +
  theme_hsci_discrete() +
  labs(fill="") +
  theme(legend.position="bottom") +
  guides(fill=guide_legend(ncol=1))

Joining with `by = join_by(clust_id)`Joining with `by = join_by(origin_work)`Joining with `by = join_by(work)`

work_verse_clusts %>% 
  inner_join(works_of_interest %>% select(work,publication_order)) %>%
  inner_join(work_verse_clusts %>% 
    inner_join(works_of_interest %>% filter(publication_order==12) %>% select(work,publication_order)), join_by(clust_id, publication_order<publication_order)
  ) %>%
  count(work.x,work.y) %>%
  inner_join(work_verse_clusts %>% count(work.x=work, name="on")) %>%
  mutate(prop=n/on) %>%
  ggplot(aes(x=work.y,y=prop,fill=work.x)) +
  geom_col(position='dodge') +
  scale_y_continuous(labels=scales::percent) +
  coord_flip() +
  theme_hsci_discrete() +
  labs(fill="") +
  theme(legend.position="bottom") +
  guides(fill=guide_legend(ncol=1))

work_verse_clusts %>% 
  inner_join(works_of_interest %>% select(work,book,publication_order)) %>%
  inner_join(work_verse_clusts %>% 
    inner_join(works_of_interest %>% filter(publication_order==12)) %>% select(clust_id,work,book,publication_order), join_by(clust_id, publication_order<publication_order)
  ) %>%
  group_by(work.x,book.y) %>%
  summarise(n=n_distinct(clust_id),.groups="drop") %>%
  inner_join(work_verse_clusts %>% count(work.x=work, name="on")) %>%
  mutate(prop=n/on) %>%
  ggplot(aes(x=work.x,y=prop)) +
  geom_col(position='dodge') +
  scale_y_continuous(labels=scales::percent) +
  coord_flip() +
  theme_hsci_discrete() +
  labs(fill="") +
  theme(legend.position="bottom") +
  guides(fill=guide_legend(ncol=1))