5 Database for Calculator Shiny
In this chapter we are going to create database, which we can use in Shiny Applications as well in results-section. In Shiny app user can add his input values for sleep apnea prevalences, condition prevalences and costs, and so get more specific results.
library(duckdb)
library(arrow)
library(dplyr)
options(scipen = 999)
## Load duckdb
con = dbConnect(duckdb::duckdb(), dbdir="data/sleep22.duckdb", read_only=TRUE)
## Load from parquet files to duckdb mem
# if(FALSE){
# con = dbConnect(duckdb::duckdb(), dbdir=":MEMORY:", read_only=FALSE)
# fils <- list.files("data/parquet_database", full.names = T)
# for (fi in fils) {
# d <- arrow::read_parquet(file = fi)
# nm <- gsub(pattern = "data/parquet_database/", replacement = "", x = fi)
# nm <- gsub(pattern = ".parquet", replacement = "", x = nm)
# duckdb::duckdb_register(conn = con, name = nm, df = d)
# }
# dbListTables(con)
# }5.1 Population datasets
Calculating 15-74 years old populations
## 15-74yrs Both, female and male population in wide format
library(tidyr)
tbl(con, "pop") %>%
filter(age_group > 14 & age_group < 75 & year_id == 2019) %>%
group_by(location_name, sex_name) %>%
summarise(pop = sum(total_population)) %>%
collect() %>%
pivot_wider(names_from = sex_name, values_from = pop) %>%
rename(pop_both = both,
pop_female = female,
pop_male = male) %>%
mutate(
pop_both = round(pop_both, 0),
pop_female = round(pop_female, 0),
pop_male = round(pop_male, 0),
ages = "15-74"
) -> pop## `summarise()` has grouped output by "location_name". You can override using the `.groups` argument.head(pop)## # A tibble: 6 × 5
## # Groups: location_name [6]
## location_name pop_both pop_female pop_male ages
## <chr> <dbl> <dbl> <dbl> <chr>
## 1 Albania 2085677 1042814 1042863 15-74
## 2 Armenia 2267463 1187751 1079711 15-74
## 3 Austria 6785345 3390812 3394534 15-74
## 4 Azerbaijan 7752896 3933115 3819781 15-74
## 5 Belarus 7249252 3828210 3421042 15-74
## 6 Belgium 8459697 4236476 4223220 15-74Total populations
## Total Both, female and male population in wide format
library(tidyr)
tbl(con, "pop") %>%
filter(year_id == 2019) %>%
group_by(location_name, sex_name) %>%
summarise(pop = sum(total_population)) %>%
collect() %>%
pivot_wider(names_from = sex_name, values_from = pop) %>%
rename(pop_both = both,
pop_female = female,
pop_male = male) %>%
mutate(
pop_both = round(pop_both, 0),
pop_female = round(pop_female, 0),
pop_male = round(pop_male, 0),
ages = "total"
) -> popu_info## `summarise()` has grouped output by "location_name". You can override using the `.groups` argument.head(popu_info)## # A tibble: 6 × 5
## # Groups: location_name [6]
## location_name pop_both pop_female pop_male ages
## <chr> <dbl> <dbl> <dbl> <chr>
## 1 Albania 2720353 1357945 1362408 total
## 2 Armenia 3019674 1562196 1457478 total
## 3 Austria 8916185 4522218 4393968 total
## 4 Azerbaijan 10278674 5136802 5141872 total
## 5 Belarus 9500785 5072640 4428145 total
## 6 Belgium 11419166 5800262 5618903 total## One dataset
pop <- pop %>% rbind(popu_info)5.2 Prevalences from ihme dataset
Calculating prevalences percentages by joining prevalences and population information
## Get cause_id's which we can use from IHME dataset
tbl(con, "causes") %>%
filter(type == "prevalence") %>%
select(cause_id) %>%
collect() -> causes
causes <- causes$cause_id[!is.na(causes$cause_id)]
## Get number of prevalences per location and causes from IHME (age: 15-74 and sex: both)
tbl(con, "ihme") %>%
filter(cause_id %in% causes & metric_name == "Number" & age_id > 7 & age_id < 20 & sex_id == 3) %>%
group_by(location_name, cause_id, cause_name) %>%
summarise(prevalence_number = sum(val)) %>%
arrange(location_name, cause_id) %>%
collect() -> prevalences_ihme## `summarise()` has grouped output by "location_name" and "cause_id". You can override using the `.groups` argument.## Calculate Prevalence Rates: join tables prevalences_ihme & pop
prevalences_ihme %>%
right_join(pop %>% filter(ages == "15-74"), by = "location_name") %>%
mutate(prevalence = prevalence_number / pop_both
) %>%
select(location_name, cause_id, cause_name, prevalence) -> prevalences_ihme
head(prevalences_ihme)## # A tibble: 6 × 4
## # Groups: location_name, cause_id [6]
## location_name cause_id cause_name prevalence
## <chr> <dbl> <chr> <dbl>
## 1 Albania 369 Maternal hypertensive disorders 0.00142
## 2 Albania 381 Neonatal preterm birth 0.00547
## 3 Albania 493 Ischemic heart disease 0.0376
## 4 Albania 494 Stroke 0.0168
## 5 Albania 536 Gastroesophageal reflux disease 0.137
## 6 Albania 594 Urinary diseases and male infertility 0.09975.3 Mortalities from ihme
Calculating mortalities for two diseases and calculating percentages by joining mortality and population information
- 1023 Other cardiovascular and circulatory diseases
- 294 All causes
## Get cause_id's which base values we can use to overwrite from IHME data
tbl(con, "causes") %>%
filter(type == "mortality") %>%
select(cause_id) %>%
collect() -> causes
causes <- causes$cause_id[!is.na(causes$cause_id)]
## Get mortality values from ihme
tbl(con, "ihme") %>%
filter(cause_id %in% causes & measure_name == "Deaths" & metric_name == "Number" & age_id > 7 & age_id < 20 & sex_id == 3) %>%
group_by(location_name, cause_id, cause_name) %>%
summarise(mortality_number = sum(val)) %>%
arrange(location_name, cause_id) %>%
collect() -> mortalities_ihme## `summarise()` has grouped output by "location_name" and "cause_id". You can override using the `.groups` argument.## Calculate Rates: join tables mortalities_ihme & pop
mortalities_ihme %>%
right_join(pop %>% filter(ages == "15-74"), by = "location_name") %>%
mutate(mortality = mortality_number / pop_both) %>%
select(location_name, cause_id, cause_name, mortality) -> mortalities_ihme
head(mortalities_ihme)## # A tibble: 6 × 4
## # Groups: location_name, cause_id [6]
## location_name cause_id cause_name mortality
## <chr> <dbl> <chr> <dbl>
## 1 Albania 294 All causes 0.00423
## 2 Albania 1023 Other cardiovascular and circulatory diseases 0.0000757
## 3 Armenia 294 All causes 0.00573
## 4 Armenia 1023 Other cardiovascular and circulatory diseases 0.0000190
## 5 Austria 294 All causes 0.00364
## 6 Austria 1023 Other cardiovascular and circulatory diseases 0.0000202## Join mortalities and prevalance data. Add name to describe value and rename 'mortality'/'prevalence' to 'value'.
prevalences_ihme$type <- "prevalence"
mortalities_ihme$type <- "mortality"
ihme <- prevalences_ihme %>%
rbind(mortalities_ihme) %>%
mutate(ihme = ifelse(is.na(prevalence), mortality, prevalence)) %>%
select(location_name, cause_id, cause_name, type, ihme) %>%
arrange(location_name, type, cause_id)
head(ihme)## # A tibble: 6 × 5
## # Groups: location_name, cause_id [6]
## location_name cause_id cause_name type ihme
## <chr> <dbl> <chr> <chr> <dbl>
## 1 Albania 294 All causes mortality 0.00423
## 2 Albania 1023 Other cardiovascular and circulatory diseases mortality 0.0000757
## 3 Albania 369 Maternal hypertensive disorders prevalence 0.00142
## 4 Albania 381 Neonatal preterm birth prevalence 0.00547
## 5 Albania 493 Ischemic heart disease prevalence 0.0376
## 6 Albania 494 Stroke prevalence 0.0168rm(list=c("mortalities_ihme", "prevalences_ihme"))5.4 Add also base values
Base data (from excel) and country specific prevalences & mortalities
## Simple causes data / PAF and base values for each country
prevalences_base_simple <- NULL
for(cntry in unique(ihme$location_name)){
tbl(con, "causes_simple") %>%
collect() %>%
mutate(location_name = cntry) %>%
select("location_name",
"cause_id",
"condition",
"PAF",
"multiplier",
"prevalence_base_italy",
"direct_healthcare_cost",
"direct_nonhealthcare_cost",
"productivity_losses_cost") %>%
rbind(prevalences_base_simple) -> prevalences_base_simple
}
head(prevalences_base_simple) #1296## # A tibble: 6 × 9
## location_name cause_id condition PAF multiplier prevalence_base_italy direct_healthcare_cost direct_nonhealthcare_cost produc…¹
## <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 United Kingdom 294 All-cause mortality 0.075 1 0.003 NA NA NA
## 2 United Kingdom 1023 Cardiovascular mortality 0.227 1 0.001 NA NA NA
## 3 United Kingdom 1029 Cancer Overall 0.097 1 0.042 5718 4581 119
## 4 United Kingdom 669 Diabetic retinopathy 0.207 0.124 0.026 307 242 579
## 5 United Kingdom 998 Diabetic kidney disease 0.135 1 0.015 797 NA NA
## 6 United Kingdom 976 Type 2 diabetes 0.145 1 0.068 3866 NA 4352
## # … with abbreviated variable name ¹productivity_losses_cost## Multirow causes data / PAF and base values for each country
prevalences_base <- NULL
for(cntry in unique(ihme$location_name)){
# print(cntry)
# tbl(con, "causes") %>%
tbl(con, "causes") %>%
collect() %>%
mutate(location_name = cntry) %>%
select("location_name",
"cause_id",
"condition",
"OSA_severity",
"gender",
"OR",
"RR",
"PAF",
"multiplier",
"prevalence_base_italy",
"direct_healthcare_cost",
"direct_nonhealthcare_cost",
"productivity_losses_cost") %>%
rbind(prevalences_base) -> prevalences_base
}
head(prevalences_base) #1296## # A tibble: 6 × 13
## location_name cause_id condition OSA_severity gender OR RR PAF multiplier prevalence_base_it…¹ direc…² direc…³ produ…⁴
## <chr> <dbl> <chr> <chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 United Kingdom 294 All-cause mortality Severe Both NA 1.54 0.075 1 0.003 NA NA NA
## 2 United Kingdom 1023 Cardiovascular mortality Severe Both NA 2.96 0.227 1 0.001 NA NA NA
## 3 United Kingdom 1029 Cancer Overall Overall Both NA 1.4 0.097 1 0.042 5718 4581 119
## 4 United Kingdom 669 Diabetic retinopathy Overall Both 2.01 NA 0.207 0.124 0.026 307 242 579
## 5 United Kingdom 998 Diabetic kidney disease Overall Both 1.59 NA 0.135 1 0.015 797 NA NA
## 6 United Kingdom 976 Type 2 diabetes Moderate-Severe Both NA 1.63 0.145 1 0.068 3866 NA 4352
## # … with abbreviated variable names ¹prevalence_base_italy, ²direct_healthcare_cost, ³direct_nonhealthcare_cost, ⁴productivity_losses_cost# 1296 *2 2592With base values, join new calculated prevalences/mortalities from IHME
## Join base data and new calculated prevalences
## Calculate condition prevalences from ihme data using multiplier (=ratio between armeni and ihme condition)
prevalences <- prevalences_base %>%
left_join(ihme, by = c("location_name", "cause_id")) %>%
select(location_name, condition, OSA_severity, gender, cause_id, cause_name, multiplier, ihme, prevalence_base_italy, OR, RR, PAF, direct_healthcare_cost, direct_nonhealthcare_cost, productivity_losses_cost) %>%
right_join(
pop %>%
filter(ages == "15-74") %>%
select(location_name, pop_both, pop_female, pop_male),
by = "location_name") %>% ## Add the population to all rows
mutate(#tamaluku = ihme,
ihme = ifelse(is.na(ihme), NA, multiplier * ihme),
multiplier = NULL)
rm(list = c("prevalences_base"))And same to simple table
## Join base data and new calculated prevalences
## Calculate condition prevalences from ihme data using multiplier (=ratio between armeni and ihme condition)
prevalences_simple <- prevalences_base_simple %>%
left_join(ihme, by = c("location_name", "cause_id")) %>%
select(location_name, condition, cause_id, cause_name, multiplier, ihme, prevalence_base_italy, PAF, direct_healthcare_cost, direct_nonhealthcare_cost, productivity_losses_cost) %>%
right_join(
pop %>%
filter(ages == "15-74") %>%
select(location_name, pop_both, pop_female, pop_male),
by = "location_name") %>% ## Add the population to all rows
mutate(ihme = ifelse(is.na(ihme), NA, multiplier * ihme),
multiplier = NULL)
head(prevalences_simple)## # A tibble: 6 × 13
## location_name condition cause_id cause_name ihme preva…¹ PAF direc…² direc…³ produ…⁴ pop_b…⁵ pop_f…⁶ pop_m…⁷
## <chr> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 United Kingdom All-cause mortality 294 All causes 4.02e-3 0.003 0.075 NA NA NA 4.99e7 2.50e7 2.48e7
## 2 United Kingdom Cardiovascular mortality 1023 Other cardiovascular and … 4.97e-5 0.001 0.227 NA NA NA 4.99e7 2.50e7 2.48e7
## 3 United Kingdom Cancer Overall 1029 Total cancers 3.80e-2 0.042 0.097 5718 4581 119 4.99e7 2.50e7 2.48e7
## 4 United Kingdom Diabetic retinopathy 669 Sense organ diseases 2.52e-2 0.026 0.207 307 242 579 4.99e7 2.50e7 2.48e7
## 5 United Kingdom Diabetic kidney disease 998 Chronic kidney disease du… 1.28e-2 0.015 0.135 797 NA NA 4.99e7 2.50e7 2.48e7
## 6 United Kingdom Type 2 diabetes 976 Diabetes mellitus type 2 1.32e-1 0.068 0.145 3866 NA 4352 4.99e7 2.50e7 2.48e7
## # … with abbreviated variable names ¹prevalence_base_italy, ²direct_healthcare_cost, ³direct_nonhealthcare_cost, ⁴productivity_losses_cost,
## # ⁵pop_both, ⁶pop_female, ⁷pop_malerm(list = c("prevalences_base_simple", "ihme"))5.6 OSA values from article (Excel)
### TODO left here, get prevelance data for slider
library(data.table)
## Read from source excel
osa <- readxl::read_xlsx("files/osa_tables.xlsx", sheet = "osa_armeni")
## Sleep apnea prevalences
osanew <- readxl::read_xlsx("files/osa_tables.xlsx", sheet = "osa_benjafield")
osanew$location_name## [1] "Afghanistan" "Albania" "Algeria" "Angola"
## [5] "Antigua and Barbuda" "Argentina" "Armenia" "Aruba"
## [9] "Australia" "Austria" "Azerbaijan" "Bahamas"
## [13] "Bahrain" "Bangladesh" "Barbados" "Belarus"
## [17] "Belgium" "Belize" "Benin" "Bhutan"
## [21] "Bolivia" "Bosnia and Herzegovina" "Botswana" "Brazil"
## [25] "Brunei" "Bulgaria" "Burkina Faso" "Burundi"
## [29] "Cape Verde" "Cambodia" "Cameroon" "Canada"
## [33] "Central African Republic" "Chad" "Chile" "China"
## [37] "Colombia" "Comoros" "Congo" "Costa Rica"
## [41] "Côte d'Ivoire" "Croatia" "Cuba" "Curacao"
## [45] "Cyprus" "Czech Republic" "Democratic Republic of the Congo" "Denmark"
## [49] "Djibouti" "Dominican Republic" "Ecuador" "Egypt"
## [53] "El Salvador" "Equatorial Guinea" "Eritrea" "Estonia"
## [57] "eSwatini" "Ethiopia" "Federated States of Micronesia" "Fiji"
## [61] "Finland" "France" "French Polynesia" "Gabon"
## [65] "Gambia" "Georgia" "Germany" "Ghana"
## [69] "Greece" "Grenada" "Guam" "Guatemala"
## [73] "Guinea" "Guinea-Bissau" "Guyana" "Haiti"
## [77] "Honduras" "Hong Kong" "Hungary" "Iceland"
## [81] "India" "Indonesia" "Iran" "Iraq"
## [85] "Ireland" "Israel" "Italy" "Jamaica"
## [89] "Japan" "Jordan" "Kazakhstan" "Kenya"
## [93] "Kiribati" "Kuwait" "Kyrgyzstan" "Laos"
## [97] "Latvia" "Lebanon" "Lesotho" "Liberia"
## [101] "Libya" "Lithuania" "Luxembourg" "Macao"
## [105] "Macedonia" "Madagascar" "Malawi" "Malaysia"
## [109] "Maldives" "Mali" "Malta" "Mauritania"
## [113] "Mauritius" "Mexico" "Moldova" "Mongolia"
## [117] "Montenegro" "Morocco" "Mozambique" "Namibia"
## [121] "Nepal" "Netherlands" "New Caledonia" "New Zealand"
## [125] "Nicaragua" "Niger" "Nigeria" "North Korea"
## [129] "Norway" "Oman" "Pakistan" "Panama"
## [133] "Papua New Guinea" "Paraguay" "Peru" "Philippines"
## [137] "Poland" "Portugal" "Puerto Rico" "Qatar"
## [141] "Romania" "Russia" "Rwanda" "Saint Lucia"
## [145] "Saint Vincent and the Grenadines" "Samoa" "São Tomé and Principe" "Saudi Arabia"
## [149] "Senegal" "Serbia" "Seychelles" "Sierra Leone"
## [153] "Singapore" "Slovakia" "Slovenia" "Solomon Islands"
## [157] "Somalia" "South Africa" "South Korea" "South Sudan"
## [161] "Spain" "Sri Lanka" "Sudan" "Suriname"
## [165] "Sweden" "Switzerland" "Syria" "Taiwan"
## [169] "Tajikistan" "Tanzania" "Thailand" "Timor Leste"
## [173] "Togo" "Tonga" "Trinidad and Tobago" "Tunisia"
## [177] "Turkey" "Turkmenistan" "Uganda" "Ukraine"
## [181] "United Arab Emirates" "United Kingdom" "USA" "Uruguay"
## [185] "Uzbekistan" "Vanuatu" "Venezuela" "Vietnam"
## [189] "Virgin Islands" "Western Sahara" "Yemen" "Zambia"
## [193] "Zimbabwe"## How many countries are needed
length(unique(prevalences$location_name))## [1] 45## Check which countries are not included in excel
length(osanew$location_name[osanew$location_name %in% unique(prevalences$location_name)])## [1] 45## 3 countries missing, check names which are missing
unique(prevalences$location_name)[!unique(prevalences$location_name) %in% osanew$location_name]## character(0)## these are not included # [1] "Andorra" "Monaco" "San Marino"
## Filter dataset and Calculate Mild, Moderate, Severe values separately
osanew <-osanew %>%
filter(location_name %in% unique(prevalences$location_name)) %>%
mutate(
Mild = `Mild-Moderate-Severe` - `Moderate-Severe`,
`Male_Moderate-Severe` = 1.3823 * `Moderate-Severe`, # Gender proportion is calculated from Italy data from Armeni article data
`Female_Moderate-Severe` = 0.6508 * `Moderate-Severe`, # Gender proportion is calculated from Italy data from Armeni article data
`Both_Moderate-Severe` = `Moderate-Severe`, # Gender proportion is calculated from Italy data from Armeni article data
Both_Mild = Mild,
Male_Mild = 0.9519 * Mild, # Gender proportion is calculated from Italy data from Armeni article data
Female_Mild = 1.0440 * Mild # Gender proportion is calculated from Italy data from Armeni article data
)
osanew <- osanew %>%
mutate(
## TODO check these with comments
Female_Moderate = 0.5342 * `Female_Moderate-Severe`, # Moderate proportion percentage is calculated from Italy data from Armeni article data
Female_Severe = 0.4658 * `Female_Moderate-Severe`, # Severe proportion percentage is calculated from Italy data from Armeni article data
Male_Moderate = 0.4004 * `Male_Moderate-Severe`, # Moderate proportion percentage is calculated from Italy data from Armeni article data
Male_Severe = 0.5996 * `Male_Moderate-Severe` ,
Both_Moderate = 0.445907 * `Both_Moderate-Severe`,
Both_Severe = 0.554093 * `Both_Moderate-Severe`
)
## To longer format
library(tidyr)
osanew <- osanew %>%
tidyr::pivot_longer(c(`Male_Moderate-Severe`, `Male_Moderate`, `Male_Severe`, Male_Mild, `Female_Moderate-Severe`, `Female_Moderate`, `Female_Severe`, Female_Mild, `Both_Moderate-Severe`, `Both_Moderate`, `Both_Severe`, Both_Mild), names_to = "OSA_severity") %>%
rename(osa_rate=value) %>%
mutate(
## Get gender and OSA_severity right
gender = sub("_.*", "", OSA_severity),
OSA_severity = sub('.+_(.+)', '\\1', OSA_severity)
) %>%
select(location_name, OSA_severity, gender, osa_rate)5.7 Money index Excel
## First try HICP index, what countries are included -----
## https://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=prc_hicp_aind&lang=en
hicp <- readxl::read_xls("files/prc_hicp_aind.xls")
hicp$`GEO/TIME`[hicp$`GEO/TIME` == "Germany (until 1990 former territory of the FRG)"] <- "Germany" ## change name of the country
## Calculate correction index to all other countries
hicp %>%
filter(`GEO/TIME` %in% unique(prevalences$location_name)) %>%
rename(location_name = `GEO/TIME`) %>%
select(location_name, `2019`) %>%
mutate(index = `2019` / hicp$`2019`[hicp$`GEO/TIME` == "Italy"],
ver = "EuroStat") %>%
select(location_name, index, ver) -> money_correction
## This is missing few countries so their index correction is 1 (equal to Italy)
## TODO find a way to correct these
# unique(prevalences_base$location_name[!prevalences_base$location_name %in% money_correction$location_name])dbDisconnect(con, shutdown=TRUE)5.8 Save data to new database
5.8.1 duckdb
# library(duckdb)
# con = dbConnect(duckdb::duckdb(), dbdir="./data/sleep22_shiny.duckdb", read_only=FALSE)
# dbWriteTable(con, "prevalences", prevalences, overwrite=TRUE)
# dbWriteTable(con, "pop", pop, overwrite=TRUE)
# dbWriteTable(con, "osa", osa, overwrite=TRUE)
# dbWriteTable(con, "osanew", osanew, overwrite=TRUE)
# dbWriteTable(con, "money_correction", money_correction, overwrite=TRUE)
# dbDisconnect(con)
#
# con = dbConnect(duckdb::duckdb(), dbdir="./data/sleep22_shiny.duckdb", read_only=FALSE)
# if(dir.exists("data/parquet_shiny/")) fs::dir_delete("data/parquet_shiny/")
# if(!dir.exists("data/parquet_shiny/")) dir.create("data/parquet_shiny/")
# dbSendQuery(con, "EXPORT DATABASE 'data/parquet_shiny/' (FORMAT PARQUET);") # copy komento? COPY taulu TO ... FORMAT
# dbDisconnect(conn = con)## Copy newdb to sleep22calculator
# fils <- c(paste0(here::here(), "/data/sleep22_shiny.duckdb"), paste0(here::here(), "/data/sleep22_shiny.duckdb.wal"))
# fs::file_copy(fils, "/Users/japmiett/projects/sleep22calculator/", overwrite = T)5.8.2 Parquet
## Testing another way to save database only to parquet files
con = dbConnect(duckdb::duckdb(), dbdir=":memory:", read_only=FALSE) ## TODO not working
## Register tbls
duckdb::duckdb_register(conn = con, name = "prevalences", df = prevalences)
duckdb::duckdb_register(conn = con, name = "prevalences_simple", df = prevalences_simple)
duckdb::duckdb_register(conn = con, name = "pop", df = pop)
duckdb::duckdb_register(conn = con, name = "osa", df = osa)
duckdb::duckdb_register(conn = con, name = "osanew", df = osanew)
duckdb::duckdb_register(conn = con, name = "money_correction", df = money_correction)
## Save to parquet
dbSendQuery(con, "COPY (SELECT * FROM prevalences) TO 'data/parquet_shiny/prevalences.parquet' (FORMAT 'parquet');")## <duckdb_result 861b0 connection=ef300 statement='COPY (SELECT * FROM prevalences) TO 'data/parquet_shiny/prevalences.parquet' (FORMAT 'parquet');'>dbSendQuery(con, "COPY (SELECT * FROM pop) TO 'data/parquet_shiny/pop.parquet' (FORMAT 'parquet');")## <duckdb_result f9890 connection=ef300 statement='COPY (SELECT * FROM pop) TO 'data/parquet_shiny/pop.parquet' (FORMAT 'parquet');'>dbSendQuery(con, "COPY (SELECT * FROM osa) TO 'data/parquet_shiny/osa.parquet' (FORMAT 'parquet');")## <duckdb_result 7af80 connection=ef300 statement='COPY (SELECT * FROM osa) TO 'data/parquet_shiny/osa.parquet' (FORMAT 'parquet');'>dbSendQuery(con, "COPY (SELECT * FROM osanew) TO 'data/parquet_shiny/osanew.parquet' (FORMAT 'parquet');")## <duckdb_result 93570 connection=ef300 statement='COPY (SELECT * FROM osanew) TO 'data/parquet_shiny/osanew.parquet' (FORMAT 'parquet');'>dbSendQuery(con, "COPY (SELECT * FROM money_correction) TO 'data/parquet_shiny/money_correction.parquet' (FORMAT 'parquet');")## <duckdb_result 7dfe0 connection=ef300 statement='COPY (SELECT * FROM money_correction) TO 'data/parquet_shiny/money_correction.parquet' (FORMAT 'parquet');'>dbSendQuery(con, "COPY (SELECT * FROM prevalences_simple) TO 'data/parquet_shiny/prevalences_simple.parquet' (FORMAT 'parquet');")## <duckdb_result 32150 connection=ef300 statement='COPY (SELECT * FROM prevalences_simple) TO 'data/parquet_shiny/prevalences_simple.parquet' (FORMAT 'parquet');'>dbDisconnect(conn = con)## Copy to sleep22calculator
fils <- list.files("data/parquet_shiny/", full.names = T)
fs::file_copy(fils, "/Users/japmiett/projects/sleep22calculator/data/", overwrite = T)