Society-aligned output analyses

knitr::opts_chunk$set(message=FALSE,warning=FALSE,dpi=300,fig.retina=2,fig.width=8)
source(here::here("src/common_basis.R"))
library(plotly)

Society-aligned output by member through time

In the following, we

1. first calculate for each member the number of society-related printings they have a substantive relationship to,
2. limit only to members for whom we have a time of admission (earliest estimate), and
3. plot each such member on the graph with x=time of their admission and y=number of society-related printings contributed to.
4. Further, we calculate a loess-smoothed average for the number of printings contributed to per member through time.

Overall

periods <- tribble(~period,~start_year,~end_year,
                   "Köthen", 1617L, 1650L,
                   "Weimar", 1651L, 1662L,
                   "Halle", 1667L, 1680L
)

p_to_a %>%
  inner_join(fbs_purpose_related_p) %>%
  inner_join(a_id_to_fbs_member_number) %>%
  filter(
    field_code %in% c("028A", "028B", "028C"),
    is.na(role) | !role %in% c("ctb", "dte"), # normed role has to be unknown or not one of these
    is.na(role2) | !str_detect(role2, !!!str_flatten(c( # role2 should not be one of these
      "^Adressat",
      "Erwähnte",
      "Gefeierte",
      "Mitglied eines Ausschusses, der akademische Grade vergibt",
      "Normerlassende Gebietskörperschaft",
      "Praeses",
      "Respondent",
      "Sonstige Person, Familie und Körperschaft",
      "Verfasser",
      "Vertragspartner",
      "Widmende",
      "Widmungsempfänger",
      "Drucker",
      "Zensor",
      "Beiträger",
      "GeistigeR Schöpfer",
      "Mitwirkender",
      "Herausgeber",
      "Angeklagte",
      "Auftraggeber"
    ), collapse = "|^"))
  ) %>%
  inner_join(fbs_metadata) %>%
  filter(field_code %in% c("028A", "028B") | is.na(rank_and_position) | !str_detect(rank_and_position,"graf|herzog|fürst")) %>%
  group_by(member_number) %>%
  summarise(works=n_distinct(p_id)) %>%
  right_join(fbs_metadata) %>%
  collect() %>%
  complete(nesting(member_number,earliest_year_of_admission),fill=list(works=0)) %>%
  inner_join(periods,join_by(earliest_year_of_admission>=start_year,earliest_year_of_admission<=end_year)) %>%
  mutate(label=str_c(member_number,": ",family_name,", ", first_name)) %>%
  ggplot(aes(x=earliest_year_of_admission,y=works)) +
  scale_x_continuous(breaks=seq(1600,1700,by=10)) +
  ylab("Mean contributions per member joining (N)") +
  xlab("Year of admission") +
  geom_smooth(span=0.3) +
  geom_point(data=. %>% group_by(period, earliest_year_of_admission) %>% summarise(works=mean(works))) +
  scale_y_continuous(breaks=seq(0,20,by=2)) +
  theme_hsci_discrete() +
  theme(legend.position = "bottom") +
  coord_cartesian(ylim=c(0,NA))

Joining with `by = join_by(p_id)`Joining with `by = join_by(a_id)`Joining with `by = join_by(member_number)`Joining with `by = join_by(member_number)`

periods <- tribble(~start_year,~end_year,~period,
1617L,1623L, "Köthen",
1624L,1630L, "Köthen",
1631L,1637L, "Köthen",
1638L,1644L, "Köthen",
1644L,1650L, "Köthen",
1651L,1656L, "Weimar",
1657L,1662L, "Weimar",
1667L,1673L, "Halle",
1674L,1680L, "Halle"
) %>% mutate(period_range=factor(str_c(start_year,"-",end_year)))

d <- p_to_a %>%
  inner_join(fbs_purpose_related_p) %>%
  inner_join(a_id_to_fbs_member_number) %>%
  filter(
    field_code %in% c("028A", "028B", "028C"),
    is.na(role) | !role %in% c("ctb", "dte"), # normed role has to be unknown or not one of these
    is.na(role2) | !str_detect(role2, !!!str_flatten(c( # role2 should not be one of these
      "^Adressat",
      "Erwähnte",
      "Gefeierte",
      "Mitglied eines Ausschusses, der akademische Grade vergibt",
      "Normerlassende Gebietskörperschaft",
      "Praeses",
      "Respondent",
      "Sonstige Person, Familie und Körperschaft",
      "Verfasser",
      "Vertragspartner",
      "Widmende",
      "Widmungsempfänger",
      "Drucker",
      "Zensor",
      "Beiträger",
      "GeistigeR Schöpfer",
      "Mitwirkender",
      "Herausgeber",
      "Angeklagte",
      "Auftraggeber"
    ), collapse = "|^"))
  ) %>%
  inner_join(fbs_metadata) %>%
  filter(field_code %in% c("028A", "028B") | is.na(rank_and_position) | !str_detect(rank_and_position,"graf|herzog|fürst")) %>%
  group_by(member_number) %>%
  summarise(works=n_distinct(p_id)) %>%
  right_join(fbs_metadata) %>%
  collect() %>%
  complete(nesting(member_number,earliest_year_of_admission),fill=list(works=0)) %>%
  inner_join(periods,join_by(earliest_year_of_admission>=start_year,earliest_year_of_admission<=end_year)) %>%
  mutate(period_range=fct_rev(period_range),printings=fct_relevel(case_when(
    works==0 ~ "0",
    works==1 ~ "1",
    works>=2 & works<5 ~ "2-4",
    works>=5 & works<10 ~ "5-9",
    works>=10 & works<20 ~ "10-19",
    works>=20 ~ ">=20"
  ), "0","1","2-4","5-9","10-19",">=20")) %>%
  mutate(label=str_c(member_number,": ",family_name,", ", first_name)) %>%
  count(period_range,printings) %>%
  group_by(period_range) %>%
  mutate(prop=n/sum(n)) %>%
  ungroup()

Joining with `by = join_by(p_id)`Joining with `by = join_by(a_id)`Joining with `by = join_by(member_number)`Joining with `by = join_by(member_number)`

d %>%
  group_by(period_range) %>%
  mutate(tn=sum(n)) %>%
  ungroup() %>%
  mutate(period_range=fct(str_c(period_range," N=(",tn,")"))) %>%
  filter(printings!="0") %>%
  ggplot(aes(x=period_range,y=prop,group=printings,fill=printings)) +
  ylab("Proportion of members") +
  xlab("Period of admission") +
  geom_col(position='stack') +
  theme_hsci_discrete() +
  theme(legend.position = "bottom") +
  scale_y_continuous(labels=scales::percent) +
  scale_coloropt(limits=c(">=20","10-19","5-9", "2-4", "1")) +
  labs(fill="Printings associated with member (N)") +
  coord_flip()

Scale for colour is already present.
Adding another scale for colour, which will replace the existing scale.

d %>%
  group_by(period_range) %>%
  mutate(tn=sum(n)) %>%
  ungroup() %>%
  mutate(period_range=fct(str_c(period_range," N=(",tn,")"))) %>%
  ggplot(aes(x=period_range,y=n,group=printings,fill=printings)) +
  ylab("Proportion of members") +
  xlab("Period of admission") +
  geom_col(position='stack') +
  theme_hsci_discrete() +
  theme(legend.position = "bottom") +
  scale_coloropt(limits=c(">=20","10-19","5-9", "2-4", "1", "0")) +
  labs(fill="Printings associated with member (N)") +
  coord_flip()

Scale for colour is already present.
Adding another scale for colour, which will replace the existing scale.

d

p1 <- p_to_a %>%
  inner_join(fbs_purpose_related_p) %>%
  inner_join(a_id_to_fbs_member_number) %>%
  filter(
    field_code %in% c("028A", "028B", "028C"),
    is.na(role) | !role %in% c("ctb", "dte"), # normed role has to be unknown or not one of these
    is.na(role2) | !str_detect(role2, !!!str_flatten(c( # role2 should not be one of these
      "^Adressat",
      "Erwähnte",
      "Gefeierte",
      "Mitglied eines Ausschusses, der akademische Grade vergibt",
      "Normerlassende Gebietskörperschaft",
      "Praeses",
      "Respondent",
      "Sonstige Person, Familie und Körperschaft",
      "Verfasser",
      "Vertragspartner",
      "Widmende",
      "Widmungsempfänger",
      "Drucker",
      "Zensor",
      "Beiträger",
      "GeistigeR Schöpfer",
      "Mitwirkender",
      "Herausgeber",
      "Angeklagte",
      "Auftraggeber"
    ), collapse = "|^"))
  ) %>% 
  group_by(member_number) %>%
  summarise(works=n_distinct(p_id)) %>%
  mutate(dataset="base set") %>%
  union_all(
    p_to_a %>%
    inner_join(fbs_purpose_related_p) %>%
    inner_join(a_id_to_fbs_member_number) %>%
    filter(
      field_code %in% c("028A", "028B", "028C"),
      is.na(role) | !role %in% c("ctb", "dte"), # normed role has to be unknown or not one of these
      is.na(role2) | !str_detect(role2, !!!str_flatten(c( # role2 should not be one of these
        "^Adressat",
        "Erwähnte",
        "Gefeierte",
        "Mitglied eines Ausschusses, der akademische Grade vergibt",
        "Normerlassende Gebietskörperschaft",
        "Praeses",
        "Respondent",
        "Sonstige Person, Familie und Körperschaft",
        "Verfasser",
        "Vertragspartner",
        "Widmende",
        "Widmungsempfänger",
        "Drucker",
        "Zensor",
        "Beiträger",
        "GeistigeR Schöpfer",
        "Mitwirkender",
        "Herausgeber",
        "Angeklagte",
        "Auftraggeber"
      ), collapse = "|^"))
    ) %>%
    inner_join(fbs_metadata) %>%
    filter(field_code %in% c("028A", "028B") | is.na(rank_and_position) | !str_detect(rank_and_position,"graf|herzog|fürst")) %>%
    group_by(member_number) %>%
    summarise(works=n_distinct(p_id)) %>%
    mutate(dataset="028C graf, herzog, fürst removed")
  ) %>%
  right_join(fbs_metadata) %>%
  collect() %>%
  complete(dataset,nesting(member_number,earliest_year_of_admission),fill=list(works=0)) %>%
  filter(!is.na(dataset)) %>%
  mutate(label=str_c(member_number,": ",family_name,", ", first_name)) %>%
  ggplot(aes(x=earliest_year_of_admission,y=works,color=dataset,fill=dataset)) +
  scale_x_continuous(breaks=seq(1600,1700,by=10)) +
  ylab("Society-related printings substantively contributed to (N)") +
  xlab("Year of admission") +
  theme_hsci_discrete()

Joining with `by = join_by(p_id)`Joining with `by = join_by(a_id)`Joining with `by = join_by(p_id)`Joining with `by = join_by(a_id)`Joining with `by = join_by(member_number)`Joining with `by = join_by(member_number)`

 
p1 + 
  geom_smooth(span=0.3) +
  geom_point(data=. %>% group_by(earliest_year_of_admission,dataset) %>% summarise(works=mean(works))) +
  scale_y_continuous(breaks=seq(0,10,by=1)) +
  theme(legend.position = "bottom") +
  coord_cartesian(ylim=c(0,NA))

(p1 + 
  scale_y_continuous(breaks=seq(0,500,by=50)) +
  geom_jitter(aes(text=label),size=0.5, height=0) +
  geom_smooth(span=0.3)
) %>%
  ggplotly(width=1024,height=768)

Warning: Ignoring unknown aesthetics: text`geom_smooth()` using method = 'loess' and formula = 'y ~ x'

periods <- tribble(~period,~start_year,~end_year,
                   "Köthen", 1617L, 1650L,
                   "Weimar", 1651L, 1667L,
                   "Halle", 1668L, 1680L
)
d <- p_to_a %>%
  inner_join(fbs_purpose_related_p) %>%
  inner_join(a_id_to_fbs_member_number) %>%
  filter(
    field_code %in% c("028A", "028B", "028C"),
    is.na(role) | !role %in% c("ctb", "dte"), # normed role has to be unknown or not one of these
    is.na(role2) | !str_detect(role2, !!!str_flatten(c( # role2 should not be one of these
      "^Adressat",
      "Erwähnte",
      "Gefeierte",
      "Mitglied eines Ausschusses, der akademische Grade vergibt",
      "Normerlassende Gebietskörperschaft",
      "Praeses",
      "Respondent",
      "Sonstige Person, Familie und Körperschaft",
      "Verfasser",
      "Vertragspartner",
      "Widmende",
      "Widmungsempfänger",
      "Drucker",
      "Zensor",
      "Beiträger",
      "GeistigeR Schöpfer",
      "Mitwirkender",
      "Herausgeber",
      "Angeklagte",
      "Auftraggeber"
    ), collapse = "|^"))
  ) %>% 
  group_by(member_number) %>%
  summarise(works=n_distinct(p_id)) %>%
  mutate(dataset="base set") %>%
  union_all(
    p_to_a %>%
    inner_join(fbs_purpose_related_p) %>%
    inner_join(a_id_to_fbs_member_number) %>%
    filter(
      field_code %in% c("028A", "028B", "028C"),
      is.na(role) | !role %in% c("ctb", "dte"), # normed role has to be unknown or not one of these
      is.na(role2) | !str_detect(role2, !!!str_flatten(c( # role2 should not be one of these
        "^Adressat",
        "Erwähnte",
        "Gefeierte",
        "Mitglied eines Ausschusses, der akademische Grade vergibt",
        "Normerlassende Gebietskörperschaft",
        "Praeses",
        "Respondent",
        "Sonstige Person, Familie und Körperschaft",
        "Verfasser",
        "Vertragspartner",
        "Widmende",
        "Widmungsempfänger",
        "Drucker",
        "Zensor",
        "Beiträger",
        "GeistigeR Schöpfer",
        "Mitwirkender",
        "Herausgeber",
        "Angeklagte",
        "Auftraggeber"
      ), collapse = "|^"))
    ) %>%
    inner_join(fbs_metadata) %>%
    filter(field_code %in% c("028A", "028B") | is.na(rank_and_position) | !str_detect(rank_and_position,"graf|herzog|fürst")) %>%
    group_by(member_number) %>%
    summarise(works=n_distinct(p_id)) %>%
    mutate(dataset="028C graf, herzog, fürst removed")
  ) %>%
  right_join(fbs_metadata) %>%
  collect() %>%
  complete(dataset,nesting(member_number,earliest_year_of_admission),fill=list(works=0)) %>%
  filter(!is.na(dataset))

Joining with `by = join_by(p_id)`Joining with `by = join_by(a_id)`Joining with `by = join_by(p_id)`Joining with `by = join_by(a_id)`Joining with `by = join_by(member_number)`Joining with `by = join_by(member_number)`

d %>% 
  inner_join(periods, join_by(earliest_year_of_admission>=start_year,earliest_year_of_admission<=end_year)) %>%
  group_by(period, dataset) %>%
  summarise(mean_printings=mean(works)) %>%
  relocate(dataset,period,mean_printings) %>%
  arrange(dataset,period)

`summarise()` has grouped output by 'period'. You can override using the `.groups` argument.

periods <- tribble(~start_year,~end_year,~period,
1617L,1623L, "Köthen",
1624L,1630L, "Köthen",
1631L,1637L, "Köthen",
1638L,1644L, "Köthen",
1644L,1650L, "Köthen",
1651L,1656L, "Weimar",
1657L,1662L, "Weimar",
1667L,1673L, "Halle",
1674L,1680L, "Halle"
) %>% mutate(period_range=factor(str_c(start_year,"-",end_year)))

p_to_a %>%
  inner_join(fbs_purpose_related_p) %>%
  inner_join(a_id_to_fbs_member_number) %>%
  filter(
    field_code %in% c("028A", "028B", "028C"),
    is.na(role) | !role %in% c("ctb", "dte"), # normed role has to be unknown or not one of these
    is.na(role2) | !str_detect(role2, !!!str_flatten(c( # role2 should not be one of these
      "^Adressat",
      "Erwähnte",
      "Gefeierte",
      "Mitglied eines Ausschusses, der akademische Grade vergibt",
      "Normerlassende Gebietskörperschaft",
      "Praeses",
      "Respondent",
      "Sonstige Person, Familie und Körperschaft",
      "Verfasser",
      "Vertragspartner",
      "Widmende",
      "Widmungsempfänger",
      "Drucker",
      "Zensor",
      "Beiträger",
      "GeistigeR Schöpfer",
      "Mitwirkender",
      "Herausgeber",
      "Angeklagte",
      "Auftraggeber"
    ), collapse = "|^"))
  ) %>% 
  inner_join(fbs_metadata) %>%
  filter(field_code %in% c("028A", "028B") | is.na(rank_and_position) | !str_detect(rank_and_position,"graf|herzog|fürst")) %>%
  group_by(member_number) %>%
  summarise(works=n_distinct(p_id)) %>%
  right_join(fbs_metadata) %>%
  replace_na(list(works=0)) %>%
  collect() %>%
  inner_join(periods, join_by(earliest_year_of_admission>=start_year,earliest_year_of_admission<=end_year)) %>%
  group_by(period_range) %>%
  summarise(published_min_one=sum(works>=1)/n(),published_min_two=sum(works>=2)/n(),published_min_five=sum(works>=5)/n(),published_min_ten=sum(works>=10)/n(),published_min_twenty=sum(works>=20)/n(), published_min_fifty=sum(works>=50)/n()) %>%
  gt() %>%
  fmt_percent() %>%
  fmt_passthrough(period_range)

Joining with `by = join_by(p_id)`Joining with `by = join_by(a_id)`Joining with `by = join_by(member_number)`Joining with `by = join_by(member_number)`

period_range	published_min_one	published_min_two	published_min_five	published_min_ten	published_min_twenty	published_min_fifty
1617-1623	13.41%	6.10%	3.66%	2.44%	1.22%	0.00%
1624-1630	6.67%	5.00%	3.33%	2.50%	0.83%	0.83%
1631-1637	9.65%	4.39%	1.75%	0.00%	0.00%	0.00%
1638-1644	9.43%	8.49%	4.72%	3.77%	2.83%	0.94%
1644-1650	19.01%	14.05%	7.44%	5.79%	4.13%	2.48%
1651-1656	15.07%	8.22%	4.79%	4.11%	2.74%	0.00%
1657-1662	20.69%	12.07%	4.31%	3.45%	2.59%	1.72%
1667-1673	32.08%	20.75%	16.98%	11.32%	3.77%	1.89%
1674-1680	30.61%	22.45%	18.37%	10.20%	4.08%	2.04%

By genre

(p_to_a %>%
  inner_join(fbs_purpose_related_p) %>%
  inner_join(a_id_to_fbs_member_number) %>%
  filter(
    field_code %in% c("028A", "028B", "028C"),
    is.na(role) | !role %in% c("ctb", "dte"), # normed role has to be unknown or not one of these
    is.na(role2) | !str_detect(role2, !!!str_flatten(c( # role2 should not be one of these
      "^Adressat",
      "Erwähnte",
      "Gefeierte",
      "Mitglied eines Ausschusses, der akademische Grade vergibt",
      "Normerlassende Gebietskörperschaft",
      "Praeses",
      "Respondent",
      "Sonstige Person, Familie und Körperschaft",
      "Verfasser",
      "Vertragspartner",
      "Widmende",
      "Widmungsempfänger",
      "Drucker",
      "Zensor",
      "Beiträger",
      "GeistigeR Schöpfer",
      "Mitwirkender",
      "Herausgeber",
      "Angeklagte",
      "Auftraggeber"
    ), collapse = "|^"))
  ) %>% 
  left_join(p_genre) %>%
  left_join(genre_categorisation) %>%
  filter(is.na(full_genre) | group_1=="Society-related") %>%
  inner_join(fbs_metadata) %>%
  filter(field_code %in% c("028A", "028B") | is.na(rank_and_position) | !str_detect(rank_and_position,"graf|herzog|fürst")) %>%
  group_by(member_number, group_3) %>%
  summarise(works=n_distinct(p_id), .groups="drop") %>%
  right_join(fbs_metadata) %>%
  mutate(label=str_c(member_number,": ",family_name,", ", first_name)) %>%
  collect() %>%
  complete(earliest_year_of_admission, group_3, fill=list(works=0)) %>%
  ggplot(aes(x=earliest_year_of_admission,y=works)) + 
  geom_jitter(aes(text=label),size=0.5, height=0) +
  geom_smooth(span=0.3) +
  scale_x_continuous(breaks=seq(1600,1700,by=10)) +
  ylab("Society-related printings substantively contributed to (N)") +
  xlab("Year of admission") +
  facet_wrap(~group_3, scales="free_y") +
  theme_hsci_discrete()) %>%
  ggplotly(width=1024,height=768)

Joining with `by = join_by(p_id)`Joining with `by = join_by(a_id)`Joining with `by = join_by(p_id)`Joining with `by = join_by(full_genre)`Joining with `by = join_by(member_number)`Joining with `by = join_by(member_number)`Warning: Ignoring unknown aesthetics: text`geom_smooth()` using method = 'loess' and formula = 'y ~ x'

Proportion of admitted members with substantive contributions to society-related printings

periods <- tribble(~start_year,~end_year,~period,
1617L,1623L, "Köthen",
1624L,1630L, "Köthen",
1631L,1637L, "Köthen",
1638L,1644L, "Köthen",
1645L,1650L, "Köthen",
1651L,1656L, "Weimar",
1657L,1662L, "Weimar",
1667L,1673L, "Halle",
1674L,1680L, "Halle"
) %>% mutate(period_range=factor(str_c(start_year,"-",end_year)))

p_to_a %>%
  inner_join(fbs_purpose_related_p) %>%
  inner_join(a_id_to_fbs_member_number) %>%
  filter(
    field_code %in% c("028A", "028B", "028C"),
    is.na(role) | !role %in% c("ctb", "dte"), # normed role has to be unknown or not one of these
    is.na(role2) | !str_detect(role2, !!!str_flatten(c( # role2 should not be one of these
      "^Adressat",
      "Erwähnte",
      "Gefeierte",
      "Mitglied eines Ausschusses, der akademische Grade vergibt",
      "Normerlassende Gebietskörperschaft",
      "Praeses",
      "Respondent",
      "Sonstige Person, Familie und Körperschaft",
      "Verfasser",
      "Vertragspartner",
      "Widmende",
      "Widmungsempfänger",
      "Drucker",
      "Zensor",
      "Beiträger",
      "GeistigeR Schöpfer",
      "Mitwirkender",
      "Herausgeber",
      "Angeklagte",
      "Auftraggeber"
    ), collapse = "|^"))
  ) %>% 
  inner_join(fbs_metadata) %>%
  filter(field_code %in% c("028A", "028B") | is.na(rank_and_position) | !str_detect(rank_and_position,"graf|herzog|fürst")) %>%
  group_by(member_number) %>%
  summarise(works=n_distinct(p_id)) %>%
  right_join(fbs_metadata) %>%
  replace_na(list(works=0)) %>%
  mutate(publishing=works>0) %>%
  collect() %>%
  inner_join(periods, join_by(earliest_year_of_admission>=start_year,earliest_year_of_admission<=end_year)) %>%
  group_by(period_range,period) %>%
  summarise(members_joined=n(),prop_publishing=sum(publishing)/n()) %>%
  ggplot(aes(x=period_range,y=prop_publishing, fill=period)) + 
  geom_col() +
  geom_text(aes(label=members_joined), nudge_y=-0.02, color="white") +
  scale_x_discrete() +
  scale_y_continuous(labels=scales::percent) +
  ylab("Proportion in publishing") +
  xlab("Period of admission") +
  theme_hsci_discrete() +
  theme(legend.position="bottom")

Joining with `by = join_by(p_id)`Joining with `by = join_by(a_id)`Joining with `by = join_by(member_number)`Joining with `by = join_by(member_number)``summarise()` has grouped output by 'period_range'. You can override using the `.groups` argument.

Numbers denote the number of members joining in each period.

Society-purpose -related publications by genre through time

Absolute

p_year %>%
  filter(year>=1600,year<1700) %>%
  inner_join(fbs_purpose_related_p) %>%
  left_join(p_genre) %>%
  left_join(genre_categorisation) %>%
  filter(is.na(full_genre) | group_1=="Society-related") %>%
  mutate(decade=floor(year/10)*10) %>%
  count(group_3,decade) %>%
  collect() %>%
  complete(group_3, decade, fill=list(n=0)) %>%
  ggplot(aes(x=decade,y=n)) +
  geom_smooth(span=0.3) +
  geom_point() +
  coord_cartesian(ylim=c(0,NA)) +
  scale_x_continuous(breaks=seq(1600,1700,by=20)) +
  ylab("Printings by members (N)") +
  xlab("Decade") +
  facet_wrap(~group_3, scales="free_y") +
  theme_hsci_discrete()

Normalised by number of members

members_by_year <- tibble(year=1617:1699) %>% 
  inner_join(fbs_metadata %>% collect(), join_by(year>=earliest_year_of_admission,year<=latest_year_of_death)) %>%
  count(year) %>%
  right_join(tibble(year=1617:1699)) %>%
  replace_na(list(n=0))

p_year %>%
  filter(year>=1600,year<1700) %>%
  inner_join(fbs_purpose_related_p) %>%
  left_join(p_genre) %>%
  left_join(genre_categorisation) %>%
  filter(is.na(full_genre) | group_1=="Society-related") %>%
  mutate(decade=floor(year/10)*10) %>%
  count(group_3,decade) %>%
  collect() %>%
  complete(group_3, decade, fill=list(n=0)) %>%
  inner_join(members_by_year %>% 
               mutate(decade=floor(year/10)*10) %>%
               group_by(decade) %>%
               summarise(mean_members=mean(n),.groups="drop")) %>%
  ggplot(aes(x=decade,y=n/mean_members)) +
  geom_smooth(span=0.3) +
  geom_point() +
  coord_cartesian(ylim=c(0,NA)) +
  scale_x_continuous(breaks=seq(1600,1700,by=20)) +
  ylab("Printings per member") +
  xlab("Decade") +
  facet_wrap(~group_3, scales="free_y") +
  theme_hsci_discrete()

Proportion of publications by genre

t <- p_year %>%
  filter(year>=1600,year<1700) %>%
  inner_join(fbs_purpose_related_p) %>%
  left_join(p_genre) %>%
  left_join(genre_categorisation) %>%
  filter(is.na(full_genre) | group_1=="Society-related") %>%
  mutate(decade=floor(year/10)*10)

t %>%
  group_by(decade) %>%
  summarise(total_n=n_distinct(p_id)) %>%
  inner_join(t %>%
    count(group_3, decade)
  ) %>%
  mutate(prop=n/total_n) %>%
  collect() %>%
  complete(group_3, decade, fill=list(prop=0)) %>%
  ggplot(aes(x=decade,y=prop)) +
  geom_smooth(span=0.3) +
  geom_point() +
  coord_cartesian(ylim=c(0,NA)) +
  scale_x_continuous(breaks=seq(1600,1700,by=20)) +
  scale_y_continuous(labels=scales::percent) +
  ylab("Society printings by genre (%)") +
  xlab("Decade") +
  facet_wrap(~group_3, scales="free_y") +
  theme_hsci_discrete()