| Type: | Package |
| Title: | Implementation of Case-Control Optimal Matching |
| Version: | 0.1.0 |
| Description: | Cases are matched to controls in an efficient, optimal and computationally flexible way. It uses the idea of sub-sampling in the level of the case, by creating pseudo-observations of controls. The user can select between replacement and without replacement, the number of controls, and several covariates to match upon. See Mamouris (2021) <doi:10.1186/s12874-021-01256-3> for an overview. |
| Depends: | R (≥ 2.10) |
| License: | GPL-2 |
| Encoding: | UTF-8 |
| LazyData: | true |
| RoxygenNote: | 7.1.1 |
| Imports: | dplyr, rlang |
| Suggests: | knitr, rmarkdown |
| VignetteBuilder: | knitr |
| NeedsCompilation: | no |
| Packaged: | 2021-04-20 10:55:41 UTC; u0112219 |
| Author: | Pavlos Mamouris [aut, cre], Vahid Nassiri [aut, ctb] |
| Maintainer: | Pavlos Mamouris <pavlos.mamouris@kuleuven.be> |
| Repository: | CRAN |
| Date/Publication: | 2021-04-21 07:40:10 UTC |
ccoptimalmatch: Optimal Case Control matching
Description
Fast and optimal matching for cases and controls
Author(s)
Maintainer: Pavlos Mamouris pavlos.mamouris@kuleuven.be
Authors:
Vahid Nassiri vahid.nassiri@openanalytics.eu [contributor]
Data for matching cases with controls
Description
A dataset containing cases and controls using the Intego registry data. The variables are as follows:
Usage
data(being_processed)
Format
A data frame with 77110 rows and 11 variables
Details
-
cluster_case: each case forms a cluster with all poosible controls to be matched
-
Patient_Id: Unique identifier for each patient
-
case_control: binary, if case==Colorectal Cancer, else control
-
case_ind: binary, if 1==case, else control
-
JCG: Year of Contact
-
entry_year: the year that the patient first entrered the database
-
CI: Comorbidity Index. Count of chronic diseases before index data
-
age_diff: difference of age between cases and controls
-
fup_diff: difference of follow-up between cases and controls
-
total_control_per_case: total controls that are available to be pooled per case
-
freq_of_controls: how many times the control is available to be matched for different cases
Not-processed data for matching cases with controls
Description
A dataset containing cases and controls using the Intego registry data. But not the final dataset. The variables are as follows:
Usage
data(not_processed)
Format
A data frame with 656506 rows and 9 variables
Details
-
Patient_Id: Unique identifier for each patient
-
JCG: Year of Contact
-
Birth_Year: Patient's year of birth
-
Gender: Patient's Gender
-
Practice_Id: Patient's general practice
-
case_control: binary, if case==Colorectal Cancer, else control
-
entry_year: the year that the patient first entrered the database
-
fup_diff: difference of follow-up between cases and controls
-
CI: Comorbidity Index. Count of chronic diseases before index data
optimal_matching
Description
optimal_matching is performing the optimal match between cases and controls in an iterative way and computational efficient way
Usage
optimal_matching(
total_database,
n_con,
cluster_var,
Id_Patient,
total_cont_per_case,
case_control,
with_replacement = FALSE
)
Arguments
total_database |
a data frame that contains the cases and controls |
n_con |
number of controls to be matched |
cluster_var |
a variable that contains one case with all available controls to be pooled |
Id_Patient |
Id of the patient |
total_cont_per_case |
total number of controls that are available for each case |
case_control |
a variable containing "case" and "control" |
with_replacement |
Use replacement or not |
Details
Here is where I should put all my details. This is where I should give more examples if necessary
Value
a data frame containing the cases and the corresponding number of controls
Examples
optimal_matching(being_processed, n_con=2, cluster_var=cluster_case,
Id_Patient=Patient_Id, total_cont_per_case=total_control_per_case, case_control = case_control)