| Title: | Cancer RADAR Project Tool |
| Version: | 4.1.0 |
| Description: | Cancer RADAR is a project which aim is to develop an infrastructure that allows quantifying the risk of cancer by migration background across Europe. This package contains a set of functions cancer registries partners should use to reshape 5 year-age group cancer incidence data into a set of summary statistics (see Boyle & Parkin (1991, ISBN:978-92-832-1195-2)) in lines with Cancer RADAR data protections rules. |
| License: | GPL (≥ 3) |
| URL: | https://cancer-radar.iarc.who.int/ |
| Depends: | R (≥ 4.1.0) |
| Imports: | dplyr (≥ 1.1.0), gtools, magrittr, openxlsx (≥ 4.2.7), purrr, rlang, rmarkdown, stats, stringr, tidyr, utils |
| Suggests: | plotly, shiny, tidyverse, DT, testthat (≥ 3.0.0), knitr, quarto |
| VignetteBuilder: | quarto |
| Config/testthat/edition: | 3 |
| Encoding: | UTF-8 |
| LazyData: | true |
| RoxygenNote: | 7.3.3 |
| NeedsCompilation: | no |
| Packaged: | 2026-10-27 15:20:06 UTC; georgesd |
| Author: | Catharina Alberts [aut, rth, pdr],
Damien Georges 
[aut, cre],
Stefano Rosso [aut],
Iacopo Baussano
[aut] |
| Maintainer: | Damien Georges <georgesd@iarc.who.int> |
| Repository: | CRAN |
| Date/Publication: | 2026-10-27 16:00:18 UTC |
Age-standardized incidence rate (asir)
Description
Age-standardized incidence rate (asir)
Usage
age_standardized_incidence_rates(
ncan,
py,
pystd,
ncan.min = 5,
py.min = 0,
ci.method = c("poisson", "legacy"),
ageg = NULL,
age.min = 0,
age.max = Inf
)
Arguments
ncan |
integer, (age-specific) number of cancers in the population of interest |
py |
integer, (age-specific) person-year in the the population of interest |
pystd |
numeric, (age-specific) standard population person-years (e.g. standard world population) |
ncan.min |
integer, minimum number of observation required not to mask the CI's out |
py.min |
integer, minimum person-years required not to mask the CI's out |
ci.method |
character, Age-standardized incidence rate (asir) and associated 95% confidence interval are computing assuming Poisson distributed counts and a normal distribution of the log(asir). asir is a summary statistics that should be computed per group of individuals providing age specific counts. The standard weights are normalised over the age groups given. |
ageg |
character, the age groups (e.g. '20_24', '85'); |
age.min |
integer, the youngest age of the window; an age group crossing |
age.max |
integer, the oldest age of the window |
Value
a 1 line and 4 column data.frame containing the asir (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
References
Boyle P, Parkin DM. Cancer registration: principles and methods. Statistical methods for registries. IARC Sci Publ. 1991;(95):126-58. PMID: 1894318.
Esteve J, Benhamou E, Raymond L. Statistical methods in cancer research. Volume IV. Descriptive epidemiology. IARC Sci Publ. 1994;(128):1-302.
Examples
ncan <- 1:10
py <- 101:110
pystd <- 10:1
ncan.min <- 5
ageg <- paste(seq(0, 45, 5), seq(4, 49, 5), sep = '_')
age_standardized_incidence_rates(ncan, py, pystd, ncan.min)
age_standardized_incidence_rates(ncan, py, pystd, ncan.min,
ageg = ageg, age.min = 20, age.max = 39)
age_standardized_incidence_rates(ncan, py, pystd, ncan.min, ci.method = 'legacy')
age_standardized_incidence_rates(ncan, py, pystd, sum(ncan) + 1)
Compute the aggregated age group names from a vector of more detailed age groups
Description
Compute the aggregated age group names from a vector of more detailed age groups
Usage
aggregated_ageg_name(selected.ageg, ageg.sep = "_")
Arguments
selected.ageg |
character, the fine grain age group vector |
ageg.sep |
character, the ageg group separator character |
Value
character, the name of the aggregated age group
Examples
ageg.in <- c('15_19', '20_24', '25_29')
aggregated_ageg_name(ageg.in)
Generate all the possible combinations of slices in a chopped vector
Description
Generate all the possible combinations of slices in a chopped vector
Usage
chop_vector(vect.size = 3)
Arguments
vect.size |
int, the size of the vector |
Value
a matrix containing all the possible slices to chop a vector per line
Examples
chop_vector(3)
Pre-computed chopped combination for vectors size 1 to 18
Description
This is a list containing all the possible combination of slices to chop vectors of size 1 to 18. It is useful to compute custom age group aggregation to ensure we are not disclosing age group with too few cancer cases.
Usage
chopped.vector.list
Format
A 18 item list:
each element is a matrix containing all the possible chop combinations to aggregate a vector of size n. ...
Compute summary statistics from 5 years age-group cancer registry data
Description
Compute summary statistics from 5 years age-group cancer registry data
Usage
create_canradar_summary_file(
filename.in,
filename.out,
ncan.min = 5,
py.min = 100,
include.by.cob.stat = TRUE,
ref = "registry",
cr.age.min = 20,
cr.age.max = 74,
verbose = TRUE
)
Arguments
filename.in |
file path, the file containing the 5 years age counts of cancers stratified per cancer type, sex and country of birth |
filename.out |
file path, the file where summary .xlsx file will be save |
ncan.min |
integer, the minimum number of cancer per age group o be displayed |
py.min |
integer, the minimum number of person-years per age group to be considered during the aggregation step (default is 100) |
include.by.cob.stat |
logical, (TRUE by default) should the statistic per country-of-birth be computed and included in the output file. |
ref |
character, the reference population(s) of the relative indices (cirr, cird, asirr, asird, crr, crd, sir, pir):
one or several of |
cr.age.min |
integer, the youngest age of the cumulative risk indices (cr, crr, crd); 20 or above, as they are computed on the aggregated age groups, which start at 20 |
cr.age.max |
integer, the oldest age of the cumulative risk indices |
verbose |
logical, shall progress message be printed |
Details
The asir leaves out the age groups without person-years or without a case count and renormalises the standard weights over the others; a warning and a line of the log sheet give the number of strata concerned.
Value
a .xlsx with all the summary statistics needed for Cancer RADAR project to be transmitted to project PIs.
Examples
## Update file.in with the path to the input file containing your registry data
## (e.g. file.filled <- "cancerRADAR_input.xlsx")
file.in <- system.file("extdata", "ex_cancerRADAR_input_filled.xlsx", package = "cancerradarr")
file.out <- 'cancerRADAR_input.xlsx'
## for cancer radar data submission, we advise to use the parameter ncan.min = 5 and
## include.by.cob.stat = TRUE
create_canradar_summary_file(file.in, file.out, ncan.min = 20, include.by.cob.stat = FALSE)
## remove the file to pass package computation tests
unlink(file.out)
Create a template file to be filled by cancer registry partners
Description
Create a template file to be filled by cancer registry partners
Usage
create_registry_input_file(filename = "cancerRADAR_input.xlsx", verbose = TRUE)
Arguments
filename |
file path, the name of the template file to be created |
verbose |
logical, shall progress message be printed |
Value
a template .xlsx file is created on the hard drive.
Examples
file.in <- 'input_file_test.xlsx'
create_registry_input_file(file.in)
## remove the file to pass package computation tests
unlink(file.in)
Create a static report from cancer RADAR output file
Description
Create a static report from cancer RADAR output file
Usage
create_static_report(filename.out = "")
Arguments
filename.out |
file path, the path to a cancer RADAR output file This function will create a html report that could be useful to check the data that will be transmitted to IARC. |
Value
nothing is returned, but a html file created with some summary statistics and graphs out of the file that should be transmitted with IARC
Cumulative risk (cr)
Description
Cumulative risk (cr)
Usage
cumulative_risk(
ncan,
py,
ageg,
age.min = 0,
age.max = 74,
ncan.min = 5,
py.min = 0
)
Arguments
ncan |
integer, (age-specific) number of cancers in the population of interest |
py |
integer, (age-specific) person-year in the the population of interest |
ageg |
character, the age groups (e.g. '20_24', '20_39', '85') |
age.min |
integer, the youngest age of the cumulative risk |
age.max |
integer, the oldest age of the cumulative risk |
ncan.min |
integer, minimum number of observation within the age window required not to mask the CI's out |
py.min |
integer, minimum person-years within the age window required not to mask the CI's out Cumulative risk (cr) between age.min and age.max and associated 95% confidence interval are computing assuming Poisson distributed counts and a normal distribution of the log cumulative rate (complementary log-log scale). An age group partly within the window contributes the years within the window only. cr is a summary statistics that should be computed per group of individuals providing age specific counts. |
Value
a 1 line and 4 column data.frame containing the cr (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
References
Esteve J, Benhamou E, Raymond L. Statistical methods in cancer research. Volume IV. Descriptive epidemiology. IARC Sci Publ. 1994;(128):1-302.
Examples
ncan <- c(2, 5, 10, 20, 30)
py <- c(5000, 5000, 5000, 7500, 5000)
ageg <- c('20_29', '30_39', '40_49', '50_64', '65_74')
cumulative_risk(ncan, py, ageg)
cumulative_risk(ncan, py, ageg, age.min = 20, age.max = 64)
cumulative_risk(ncan, py, ageg, ncan.min = sum(ncan) + 1)
Cumulative risk difference (crd)
Description
Cumulative risk difference (crd)
Usage
cumulative_risk_difference(
ncan,
py,
ncanref,
pyref,
ageg,
ageg.ref = ageg,
age.min = 0,
age.max = 74,
ncan.min = 5,
py.min = 0
)
Arguments
ncan |
integer, (age-specific) number of cancers in the population of interest |
py |
integer, (age-specific) person-year in the the population of interest |
ncanref |
integer, (age-specific) number of cancers in the reference population |
pyref |
integer, (age-specific) person-year in the the reference population |
ageg |
character, the age groups (e.g. '20_24', '20_39', '85') of the population of interest |
ageg.ref |
character, the age groups of the reference population (the same as ageg by default) |
age.min |
integer, the youngest age of the cumulative risk |
age.max |
integer, the oldest age of the cumulative risk |
ncan.min |
integer, minimum number of observation within the age window required not to mask the CI's out |
py.min |
integer, minimum person-years within the age window required not to mask the CI's out Cumulative risk difference (crd) between age.min and age.max and associated 95% confidence interval are computing assuming Poisson distributed counts and a normal distribution of the crd. crd is a summary statistics that should be computed per group of individuals providing age specific counts. |
Value
a 1 line and 4 column data.frame containing the crd (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
References
Esteve J, Benhamou E, Raymond L. Statistical methods in cancer research. Volume IV. Descriptive epidemiology. IARC Sci Publ. 1994;(128):1-302.
Examples
ncan <- c(2, 5, 10, 20, 30)
py <- c(5000, 5000, 5000, 7500, 5000)
ncanref <- c(20, 60, 150, 400, 500)
pyref <- c(50000, 60000, 70000, 90000, 60000)
ageg <- c('20_29', '30_39', '40_49', '50_64', '65_74')
cumulative_risk_difference(ncan, py, ncanref, pyref, ageg)
Cumulative risk ratio (crr)
Description
Cumulative risk ratio (crr)
Usage
cumulative_risk_ratio(
ncan,
py,
ncanref,
pyref,
ageg,
ageg.ref = ageg,
age.min = 0,
age.max = 74,
ncan.min = 5,
py.min = 0
)
Arguments
ncan |
integer, (age-specific) number of cancers in the population of interest |
py |
integer, (age-specific) person-year in the the population of interest |
ncanref |
integer, (age-specific) number of cancers in the reference population |
pyref |
integer, (age-specific) person-year in the the reference population |
ageg |
character, the age groups (e.g. '20_24', '20_39', '85') of the population of interest |
ageg.ref |
character, the age groups of the reference population (the same as ageg by default) |
age.min |
integer, the youngest age of the cumulative risk |
age.max |
integer, the oldest age of the cumulative risk |
ncan.min |
integer, minimum number of observation within the age window required not to mask the CI's out |
py.min |
integer, minimum person-years within the age window required not to mask the CI's out Cumulative risk ratio (crr) between age.min and age.max and associated 95% confidence interval are computing assuming Poisson distributed counts and a normal distribution of the log(crr). crr is a summary statistics that should be computed per group of individuals providing age specific counts. |
Value
a 1 line and 4 column data.frame containing the crr (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
References
Esteve J, Benhamou E, Raymond L. Statistical methods in cancer research. Volume IV. Descriptive epidemiology. IARC Sci Publ. 1994;(128):1-302.
Examples
ncan <- c(2, 5, 10, 20, 30)
py <- c(5000, 5000, 5000, 7500, 5000)
ncanref <- c(20, 60, 150, 400, 500)
pyref <- c(50000, 60000, 70000, 90000, 60000)
ageg <- c('20_29', '30_39', '40_49', '50_64', '65_74')
cumulative_risk_ratio(ncan, py, ncanref, pyref, ageg)
Smart aggregation of cancer cases per age group
Description
Smart aggregation of cancer cases per age group
Usage
custom_ageg_aggregation(
dat,
ncan.min = 5,
py.min = 0,
add.total = FALSE,
ncan.lab = "ncan",
py.lab = "py"
)
Arguments
dat |
tibble, a single cancer/sex/country tibble containing cancer cases from a registry. It should contains the column ageg and ncan |
ncan.min |
integer, the minimal number of cancer in each category |
py.min |
integer, the minimal number of person-years in each category |
add.total |
logical, should the 'total' category added to the output dataset |
ncan.lab |
character, the column label where cancer cases are stored |
py.lab |
character, the column label where (optional) population at risk are stored |
Value
aggregated dataset where all the age group contains at least ncan.min cancers cases; a stratum with fewer than ncan.min cases (or py.min person-years) in total is returned as a single age group
Examples
dat <-
dplyr::tribble(
~ ageg, ~ ncan,
'00_04', 0,
'05_09', 0,
'10_14', 0,
'15_19', 0,
'20_24', 1,
'25_29', 2,
'30_34', 4,
'35_39', 5,
'40_44', 1,
'45_49', 10,
'50_54', 14,
'55_59', 1,
'60_64', 2,
'65_69', 2,
'70_74', 5,
'75_79', 1,
'80_84', 0,
'85', 0
)
custom_ageg_aggregation(dat, 0)
custom_ageg_aggregation(dat, 5)
custom_ageg_aggregation(dat, 10)
custom_ageg_aggregation(dat, 100)
Geographical aggregation used for cancerradarr
Description
In order to prevent loose of data in case of too low effective, several geographical aggregation can be considered. In this table are stored the different level of aggregation and the aggregation correspondence table considered.
Usage
dat.aggr
Format
A data frame with 250 rows and 5 columns:
- cob_iso3
Country ISO3 code
- un_region
UN region
- un_subregion
UN subregion
- hdi_cat
HDI 2023 category
- any_migr
any migration background
...
Details
A multi-columns dataset containing all the countries of birth (as ISO3 code) and other geographical aggregation rules
Burden of cancer aggregation category used for cancerradarr
Description
A multi-columns dataset containing for all the countries of birth (as ISO3 code), sex and cancer type
combinations the quartiles of cancer burden in country of origin burden. The quartiles (0%-24%,
25%-49%, 50%-74% and 75-100%) are based on the ASIR from GLOBOCAN 2022.
Usage
dat.asr.cat
Format
A data frame with 2,220 rows and 5 columns:
- cob_iso3
Country ISO3 code
- sex
targeted sex
- can
the cancer type
- asr
GLOBOCAN 2022 age-standardized cancer incidence rate
- asr_rank_cat
GLOBOCAN 2022 age-standardized cancer incidence rate quartile category
...
Countries label and countries codes
Description
A 2 column dataset containing all the countries of birth (with associated countries codes) included in Cancer RADAR project
Usage
dat.cob
Format
A data frame with 251 rows and 3 columns:
- cob_label
Country name
- cob_code
Country code
- cob_iso3
Country ISO3 code (used as unique id)
...
Diagnose Input File for Data Quality Issues
Description
This function checks the input file for common data quality issues that can cause errors during processing, including column names that differ from the template, implausible data, duplicate age groups and unexpected age group labels.
Usage
diagnose_input_file(filename.in)
Arguments
filename.in |
file path, the file containing the 5 years age counts of cancers stratified per cancer type, sex and country of birth |
Value
A list containing diagnostic information:
column_issues: Data frame of the columns that differ from the template
data_checks: Data frame of the data checks that failed (check, severity, sheet, detail)
duplicate_age_groups: Data frame of combinations with duplicate age groups
invalid_age_groups: Character vector of invalid age group labels found
summary: Overall summary of issues found
Examples
file.in <- system.file("extdata", "ex_cancerRADAR_input_filled.xlsx",
package = "cancerradarr")
diagnosis <- diagnose_input_file(file.in)
print(diagnosis$summary)
European countries age-specific cancer burden from GLOBOCAN 2022
Description
A multi-columns dataset containing for all the European countries (UN definition) (as ISO3 code), sex and cancer type
combinations the number of cases and population at risk estimated in GLOBOCAN 2022. This data
are used in cancerradarr to compute the relative index on a standard reference population that
could be more easily compared between registries.
In addition to individual European countries, aggregated areas such as
E27 (European Union 27 countries) and EUN (all the UN European countries)
are stred in the dataset
Colorectal cancer (colo) is GLOBOCAN 'Colon' (C18) + 'Rectum' (C19-C20).
Up to version 3.0.0 it was 'Colon' alone. The dataset is built by
data-raw/globocan_2022_eu.R.
Usage
globocan.2022.eu
Format
A data frame with 9,576 rows and 6 columns:
- cob_iso3
Country ISO3 code
- sex
targeted sex
- ageg
targeted age group
- can
the cancer type
- ncanref
number of cancer cases estimated in GLOBOCAN 2022
- pyref
population at risk estimated in GLOBOCAN 2022
...
Source
https://gco.iarc.fr/today/en
References
Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, Jemal A. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024 May-Jun;74(3):229-263. doi: 10.3322/caac.21834. Epub 2024 Apr 4. PMID: 38572751.
Compute crude incidence rates
Description
Compute crude incidence rates
Usage
incidence_rates(ncan, py, ncan.min = 5, py.min = 0)
Arguments
ncan |
integer, number of cancer |
py |
integer, number of person-year |
ncan.min |
integer, minimum number of observation required not to mask the CI's out |
py.min |
integer, minimum number of person-year required not to mask the CI's out Crude incidence rates and associated 95% confidence interval are computing assuming a Poisson distribution and the exact method. |
Value
a 4 column data.frame containing the crude incidence rate estimate (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
References
Boyle P, Parkin DM. Cancer registration: principles and methods. Statistical methods for registries. IARC Sci Publ. 1991;(95):126-58. PMID: 1894318.
Examples
ncan <- c(1, 10, 100)
py <- c(10, 100, 1000)
incidence_rates(ncan, py, 5, 500)
Compute incidence rates difference
Description
Compute incidence rates difference
Usage
incidence_rates_difference(ncan, py, ncanref, pyref, ncan.min = 5, py.min = 0)
Arguments
ncan |
integer, number of cancers in the population of interest |
py |
integer, person-year of the the population of interest |
ncanref |
integer, number of cancers in the reference population |
pyref |
integer, person-year of the the reference population |
ncan.min |
integer, minimum number of observation required not to mask the CI's out |
py.min |
integer, minimum person-year required not to mask the CI's out Incidence rates differences and associated 95% confidence interval are computing assuming normal distribution of the differences.. |
Value
a 4 column data.frame containing the incidence rates difference (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
Examples
ncan <- 1:10
py <- 101:110
ncanref <- 41:50
pyref <- 251:260
ncan.min <- 5
incidence_rates_difference(ncan, py, ncanref, pyref, ncan.min)
Compute incidence rates ratio
Description
Compute incidence rates ratio
Usage
incidence_rates_ratio(ncan, py, ncanref, pyref, ncan.min = 5, py.min = 0)
Arguments
ncan |
integer, number of cancers in the population of interest |
py |
integer, person-year of the the population of interest |
ncanref |
integer, number of cancers in the reference population |
pyref |
integer, person-year of the the reference population |
ncan.min |
integer, minimum number of observation required not to mask the CI's out |
py.min |
integer, minimum person-year required not to mask the CI's out Incidence rates ratio and associated 95% confidence interval are computing assuming normal distribution of the ratios on the log scale. |
Value
a 4 column data.frame containing the incidence rates ratio (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
References
Boyle P, Parkin DM. Cancer registration: principles and methods. Statistical methods for registries. IARC Sci Publ. 1991;(95):126-58. PMID: 1894318.
Examples
ncan <- 1:10
py <- 101:110
ncanref <- 41:50
pyref <- 251:260
ncan.min <- 5
incidence_rates_ratio(ncan, py, ncanref, pyref, ncan.min)
Compute the indirect proportional incidence ratio (pir)
Description
Compute the indirect proportional incidence ratio (pir)
Usage
indirect_proportional_incidence_ratio(
ncan,
ntot,
ncanref,
ntotref,
ncan.min = 5
)
Arguments
ncan |
integer, (age-specific) number of cancers in the population of interest |
ntot |
integer, (age-specific) total number of cancer the the population of interest |
ncanref |
integer, (age-specific) number of cancers in the reference population |
ntotref |
integer, (age-specific) total number of cancer the the reference of interest |
ncan.min |
integer, minimum number of observation required not to mask the CI's out Indirect proportional incidence ratio and associated 95% confidence interval are computing assuming normal distribution of the pir on the log scale. pir is a summary statistics that should be computed per group of individuals providing age specific counts. |
Value
a 1 line and 4 column data.frame containing the pir (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
References
Boyle P, Parkin DM. Cancer registration: principles and methods. Statistical methods for registries. IARC Sci Publ. 1991;(95):126-58. PMID: 1894318.
Examples
ncan <- 1:10
ntot <- 11:20
ncanref <- 41:50
ntotref <- 251:260
ncan.min <- 5
indirect_proportional_incidence_ratio(ncan, ntot, ncanref, ntotref, ncan.min)
indirect_proportional_incidence_ratio(ncan, ntot, ncanref, ntotref, sum(ncan) + 1)
Compute indirect standardized incidence ratio (sir)
Description
Compute indirect standardized incidence ratio (sir)
Usage
indirect_standardized_incidence_ratio(
ncan,
py,
ncanref,
pyref,
ncan.min = 5,
py.min = 0
)
Arguments
ncan |
integer, (age-specific) number of cancers in the population of interest |
py |
integer, (age-specific) person-year of the the population of interest |
ncanref |
integer, (age-specific) number of cancers in the reference population |
pyref |
integer, (age-specific) person-year of the the reference population |
ncan.min |
integer, minimum number of observation required not to mask the CI's out |
py.min |
integer, minimum number of person-years required not to mask the CI's out Standardized incidence ratio (sir) and associated 95% confidence interval are computing assuming normal distribution of the pir on the log scale. sir is a summary statistics that should be computed per group of individuals providing age specific counts. |
Value
a 1 line and 4 column data.frame containing the sir (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
References
Boyle P, Parkin DM. Cancer registration: principles and methods. Statistical methods for registries. IARC Sci Publ. 1991;(95):126-58. PMID: 1894318.
Examples
ncan <- 1:10
py <- 101:110
ncanref <- 41:50
pyref <- 251:260
ncan.min <- 5
indirect_standardized_incidence_ratio(ncan, py, ncanref, pyref, ncan.min)
indirect_standardized_incidence_ratio(ncan, py, ncanref, pyref, sum(ncan) + 1)
Open cancer RADAR output file dictionary
Description
Calling this function will open the dictionary describing sheets and variables stored in the cancer
summary file (output file generated by create_canradar_summary_file function). It
could be useful for the cancer registries to check what kind of data they will be sharing.
Note that a temporary copy of the dictionary is created on your hard drive to prevent from
unwanted file modification.
Usage
open_canradar_dictionary()
Value
the path to a temporary file where cancer RADAR dictionary is stored
Examples
## Not run:
open_canradar_dictionary()
## End(Not run)
Compute proportional rates
Description
Compute proportional rates
Usage
proportional_rates(ncan, ntot, ncan.min = 5)
Arguments
ncan |
integer, number of cancer of interest |
ntot |
integer, overall number of cancer |
ncan.min |
integer, minimum number of observation required not to mask the CI's out Proportional incidence rates and associated 95% confidence interval are computing assuming a Binomial distribution and the Clopper and Pearson (1934) procedure. |
Value
a 4 column data.frame containing the proportional incidence rate estimate (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
References
C. J. CLOPPER, B.Sc., E. S. PEARSON, D.Sc., THE USE OF CONFIDENCE OR FIDUCIAL LIMITS ILLUSTRATED IN THE CASE OF THE BINOMIAL, Biometrika, Volume 26, Issue 4, December 1934, Pages 404–413, https://doi.org/10.1093/biomet/26.4.404
Boyle P, Parkin DM. Cancer registration: principles and methods. Statistical methods for registries. IARC Sci Publ. 1991;(95):126-58. PMID: 1894318.
See Also
Examples
ncan <- c(1, 10, 100)
ntot <- c(10, 100, 1000)
proportional_rates(ncan, ntot, 5)
Read the data information of a cancer RADAR file
Description
Read the data information of a cancer RADAR file
Usage
read_cancerradar_data_info(filename)
Arguments
filename |
file path, a cancer RADAR output file (or input file) |
Value
a 1 line tibble with 13 columns
country: character, the registry country
registry: character, the cancer registry designation
period: character, the period over which the data were collected, as written
period_start, period_end: integer, its first and last year (NA if not two 4-digit years)
source_cob: character, the data source of the country of birth
source_py: character, the data source of the population at risk
version: character, the cancerradarr version that created the file
ncan_min: dbl, the number of cases threshold
py_min: dbl, the number of person-years threshold
references: character, the reference populations of the relative indices
cr_window: character, the age window of the cumulative risk indices (e.g. '20-74')
created: character, the date the file was created
The last six come from the log sheet of output files; they are NA for an input file, or when an older file did not record them.
Examples
filename.out <- system.file('extdata/ex_cancerRADAR_output.xlsx', package = "cancerradarr")
read_cancerradar_data_info(filename.out)
Read cancer registry summary statistics (non age-specific)
Description
Read cancer registry summary statistics (non age-specific)
Usage
read_cancerradar_output_01(filename.out, aggr.level = "cob_iso3")
Arguments
filename.out |
file path, the path to a cancer RADAR output file |
aggr.level |
character, the aggregation level to be considered. Should be one of |
Value
a tibble with 10 columns
reg_label: factor, the chosen aggregation level id
sex: character, male/female
ageg: character, age group (here
total)can: character, the cancer type
ref: character, the reference population for relative index
index: character, the type of index
est: dbl, the index estimator
lci: dbl, the index confidence interval lower bound
uci: dbl, the index confidence interval upper bound
censored: lgl, TRUE when the value is censored (below ncan.min / py.min) rather than missing
Examples
filename.out <- system.file('extdata/ex_cancerRADAR_output.xlsx', package = "cancerradarr")
dat.out <- read_cancerradar_output_01(filename.out, 'un_region')
head(dat.out)
Read cancer registry summary statistics (age-specific incidence rate and proportional rates)
Description
Read cancer registry summary statistics (age-specific incidence rate and proportional rates)
Usage
read_cancerradar_output_02(filename.out, aggr.level = "cob_iso3")
Arguments
filename.out |
file path, the path to a cancer RADAR output file |
aggr.level |
character, the aggregation level to be considered. Should be one of |
Value
a tibble with 12 columns
reg_label: factor, the chosen aggregation level id
sex: character, male/female
ageg: character, age group (here
total)can: character, the cancer type
index: character, the type of index
est: dbl, the index estimator
lci: dbl, the index confidence interval lower bound
uci: dbl, the index confidence interval upper bound
censored: lgl, TRUE when the value is censored (below ncan.min / py.min) rather than missing
ageg_sta: dbl, the age group starting age
ageg_sto: dbl, the age group stopping age
ageg_mid: dbl, the age group middle age
Examples
filename.out <- system.file('extdata/ex_cancerRADAR_output.xlsx', package = "cancerradarr")
dat.out <- read_cancerradar_output_02(filename.out, 'un_region')
head(dat.out)
Read cancer registry summary statistics (age-specific number of cancer cases)
Description
Read cancer registry summary statistics (age-specific number of cancer cases)
Usage
read_cancerradar_output_03(filename.out, aggr.level = "cob_iso3")
Arguments
filename.out |
file path, the path to a cancer RADAR output file |
aggr.level |
character, the aggregation level to be considered. Should be one of |
Value
a tibble with 10 columns
reg_label: factor, the chosen aggregation level id
sex: character, male/female
ageg: character, age group (here
total)can: character, the cancer type
ncan: dbl, the number of cancer
py: dbl, the population size (if available)
censored: lgl, TRUE when the number of cancer is censored (below ncan.min) rather than missing
ageg_sta: dbl, the age group starting age
ageg_sto: dbl, the age group stopping age
ageg_mid: dbl, the age group middle age
Examples
filename.out <- system.file('extdata/ex_cancerRADAR_output.xlsx', package = "cancerradarr")
dat.out <- read_cancerradar_output_03(filename.out, 'any_migr')
head(dat.out)
Create a dynamic report from cancer RADAR output file
Description
Create a dynamic report from cancer RADAR output file
Usage
run_dynamic_report(filename.out = "")
Arguments
filename.out |
file path, the path to a cancer RADAR output file This function will open a shiny app where cancer registries can visually check the data they will be transmitted to IARC. |
Value
nothing is returned
Age-standardized incidence rates differences (asird)
Description
Age-standardized incidence rates differences (asird)
Usage
standardized_incidence_rate_difference(
ncan,
py,
ncanref,
pyref,
pystd,
ncan.min = 5,
py.min = 0,
ci.method = c("poisson", "legacy"),
ageg = NULL,
age.min = 0,
age.max = Inf
)
Arguments
ncan |
integer, (age-specific) number of cancers in the population of interest |
py |
integer, (age-specific) person-year in the the population of interest |
ncanref |
integer, (age-specific) number of cancers in the reference population |
pyref |
integer, (age-specific) person-year in the the reference population |
pystd |
numeric, (age-specific) standard population person-years (e.g. standard world population) |
ncan.min |
integer, minimum number of observation required not to mask the CI's out |
py.min |
integer, minimum number of person-years required not to mask the CI's out |
ci.method |
character, Age-standardized incidence rate difference (asird) and associated 95% confidence interval are computing assuming Poisson distributed counts and a normal distribution of the asird. asird is a summary statistics that should be computed per group of individuals providing age specific counts. |
ageg |
character, the age groups (e.g. '20_24', '85'); |
age.min |
integer, the youngest age of the window; an age group crossing |
age.max |
integer, the oldest age of the window |
Value
a 1 line and 4 column data.frame containing the pir (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
References
https://www.hsph.harvard.edu/thegeocodingproject/analytic-methods/
Examples
ncan <- 1:10
py <- 101:110
ncanref <- 41:50
pyref <- 251:260
pystd <- 10:1
ncan.min <- 5
ageg <- paste(seq(0, 45, 5), seq(4, 49, 5), sep = '_')
standardized_incidence_rate_difference(ncan, py, ncanref, pyref, pystd, ncan.min)
standardized_incidence_rate_difference(ncan, py, ncanref, pyref, pystd, ncan.min,
ageg = ageg, age.min = 20, age.max = 39)
standardized_incidence_rate_difference(ncan, py, ncanref, pyref, pystd, sum(ncan) + 1)
Age-standardized incidence rates ratio (asirr)
Description
Age-standardized incidence rates ratio (asirr)
Usage
standardized_incidence_rate_ratio(
ncan,
py,
ncanref,
pyref,
pystd,
ncan.min = 5,
py.min = 0,
ci.method = c("poisson", "legacy"),
ageg = NULL,
age.min = 0,
age.max = Inf
)
Arguments
ncan |
integer, (age-specific) number of cancers in the population of interest |
py |
integer, (age-specific) person-year in the the population of interest |
ncanref |
integer, (age-specific) number of cancers in the reference population |
pyref |
integer, (age-specific) person-year in the the reference population |
pystd |
numeric, (age-specific) standard population person-years (e.g. standard world population) |
ncan.min |
integer, minimum number of observation required not to mask the CI's out |
py.min |
integer, minimum person-year required not to mask the CI's out |
ci.method |
character, Age-standardized incidence rate ratio (asirr) and associated 95% confidence interval are computing assuming Poisson distributed counts and a normal distribution of the log(asirr). asirr is a summary statistics that should be computed per group of individuals providing age specific counts. |
ageg |
character, the age groups (e.g. '20_24', '85'); |
age.min |
integer, the youngest age of the window; an age group crossing |
age.max |
integer, the oldest age of the window |
Value
a 1 line and 4 column data.frame containing the pir (est) and associated 95% CI (lci, uci) and censored (TRUE when the CI is masked because of ncan.min or py.min)
References
Boyle P, Parkin DM. Cancer registration: principles and methods. Statistical methods for registries. IARC Sci Publ. 1991;(95):126-58. PMID: 1894318.
Examples
ncan <- 1:10
py <- 101:110
ncanref <- 41:50
pyref <- 251:260
pystd <- 10:1
ncan.min <- 5
ageg <- paste(seq(0, 45, 5), seq(4, 49, 5), sep = '_')
standardized_incidence_rate_ratio(ncan, py, ncanref, pyref, pystd, ncan.min)
standardized_incidence_rate_ratio(ncan, py, ncanref, pyref, pystd, ncan.min,
ageg = ageg, age.min = 20, age.max = 39)
standardized_incidence_rate_ratio(ncan, py, ncanref, pyref, pystd, ncan.min, ci.method = 'legacy')
standardized_incidence_rate_ratio(ncan, py, ncanref, pyref, pystd, sum(ncan) + 1)
Validate Age Group Labels
Description
Checks that age group labels in a dataset match expected formats.
Usage
validate_age_groups(dat)
Arguments
dat |
A data frame containing an |
Details
Expected age group labels are:
Five-year age bands: '00_04', '05_09', ..., '80_84'
Open-ended group: '85'
Special categories: 'total', 'DCO', 'MV'
Value
NULL (invisibly). The function is called for its side effect of stopping execution if invalid age groups are found.
Examples
dat <- data.frame(ageg = c("00_04", "05_09", "10_14"))
validate_age_groups(dat)