A vector containing a covariate to match the conditions on.

A factor vector containing condition labels.

A vector containing a covariate to match the conditions on.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

The criterion function to use, such as .t_crit.

A factor vector containing condition labels.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

A columnwise matrix containing covariates to match the conditions on.

The criterion function to use, such as .t_crit.

A factor vector containing condition labels.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

The number of subjects for each group.

The number of subjects that can be removed for each group.

The expected proportion of subjects for each group.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

If FALSE, prints warning when the test statistic cannot be calculated; if TRUE (the default) they are not printed.

The number of different condition values, usually created by calling table(condition).

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

The best matched groups so far together with its p-value / thresh ratio; a list containing ratio and sets (a list of subject index vectors).

A set of group sizes as a data.table row (also a list).

An ordered subject subspace: a list of vectors, with one vector per group containing the corresponding subject indices.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

If TRUE, prints summary information on the input and the results, as well as progress information for the exhaustive search and random algorithms. Default: TRUE; can be changed using set_param("PRINT_INFO", FALSE).

An iterator returning (or a list containing) indices for the is.in logical vector whose in / out status is to be changed.

A logical vector showing which items are preserved currently; versions resulting by changing indices for each candidate are then compared.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

If TRUE, prefers higher test statistic more than the expected group size proportion; default is TRUE. Used by all algorithms except exhaustive, which always

A named integer vector, containing the maximum number of subjects that can be removed from each group. Specify 0 for groups if you want to preserve all of their subjects. If you do not specify a value for a group, it defaults to 2 less than the group size. Values outside the valid range of 0..(N-1) (where N is the number of subjects in the group) are corrected without a warning.

The p-value / threshold ratio at which it starts removing subjects one by one. Used when max_removed_per_step > 1, with a default value of 0.5.

A data.table with an ind column with indices for items to consider dropping.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

A list(metric, sets); metric is a number, set is a list of vectors.

A list(metric, set); metric is a number, set is a vector. Candidate is only considered if metric is not zero.

A list of initializer functions (with no arguments) for iterators.

A function for retrieving next item for an iterator argument; it assumes that the iterator returns NULL when finished.

elements to be used (a list of vectors)

A data.table with the columns containing the group names, and the rows containing a particular setup of group sizes. All rows are expected to have the same sum (not checked).

The group names (specified because the table can have other columns as well).

The minimum number of subjects to be preserved per group.

A vector containing a covariate to match the conditions on.

A factor vector containing condition labels.

Integer indices for the is.in logical vector.

A logical vector showing which items are preserved.

An iterator created using either the iterpc or the iterators package, or anything else foreach::foreach can interpret (esp. a list).

The operation to be performed for each item in input (possibly after preprocessing it; see preprocess_input).

Processes each value retrieved from the input iterator.

The same as the parameters of foreach::foreach with identical names.

The maximum number of items to be retrieved from input if it is an iterator.

If TRUE, prints messages about the progress.

Used to use iterpc::iter_wrapper() on iterpc iterators, but realized that foreach doesn't handle iterators in a nice way (converts it to a list, which may be huge, instead of gradually retrieving the contensts), so feeding segments of the iterators to foreach instead.

The name of one halting test, or a halting test function.

The number of seconds to convert to human-readable form.

The number of decimals to print in the output.

A function; default: the caller.

The parent frame index; default: 3 (the great-grandparent).

A logical vector for output 1, TRUE iff row is in the match.

A logical vector for output 2, TRUE iff row is in the match.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

If TRUE, it prioritizes the test statistic more than the group size proportion.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

A vector containing a covariate to match the conditions on.

A factor vector containing condition labels.

A vector containing a covariate to match the conditions on.

A factor vector containing condition labels.

A named integer vector, containing the maximum number of subjects that can be removed from each group. Specify 0 for groups if you want to preserve all of their subjects. If you do not specify a value for a group, it defaults to 2 less than the group size. Values outside the valid range of 0..(N-1) (where N is the number of subjects in the group) are corrected without a warning.

A factor vector containing condition labels.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

A factor vector containing condition labels.

If TRUE and props is a character vector, last item is not dropped.

Threshold value(s).

Halting tests.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

The maximum number of subjects that can be removed from each group. It must have a valid number for each group.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

The minimum number of preserved subjects. It can be used to ensure that the search will not take forever to run, but instead fail when a solution is not found when preserving this number of subjects.

The lookahead to use: a positive integer. It is used by the heuristic3 and heuristic4 algorithms, with a default of 2. The running time is O(N ^ lookahead), wheren N is the number of subjects.

If TRUE, prefers higher test statistic more than the expected group size proportion; default is TRUE. Used by all algorithms except exhaustive, which always

If TRUE, prints summary information on the input and the results, as well as progress information for the exhaustive search and random algorithms. Default: TRUE; can be changed using set_param("PRINT_INFO", FALSE).

The number of equivalent subjects that can be removed in each step. (The actual allowed number may be less depending on the p-value / theshold ratio.) This parameters is used by the heuristic3 and heuristic4 algorithms, with a default value of 1.

The percentage of remaining subjects that can be removed in each step. Used when max_removed_per_step > 1, with a default value of 0.5.

The p-value / threshold ratio at which it starts removing subjects one by one. Used when max_removed_per_step > 1, with a default value of 0.5.

The names of arguments given to the search function.

Consumes extra parameters that are not used by the search algorithm at hand; this function gives a warning about the ones whose value is not NULL that their value is not used.

A data.table with the columns containing the group names, and the rows containing a particular setup of group sizes. All rows are expected to have the same sum (not checked).

The group names (specified because the table can have other columns as well).

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

A vector containing a covariate to match the conditions on.

A factor vector containing condition labels.

A list.

A list of vectors.

A string to be inserted between the name of a vector item and its value.

The names of arguments given to the search function.

Consumes extra parameters that are not used by the search algorithm at hand; this function gives a warning about the ones whose value is not NULL that their value is not used.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

The output of match_groups(): either a logical vector, or a list of those.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

A logical vector for output 1, TRUE iff row is in the match.

A logical vector for output 2, TRUE iff row is in the match.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

If TRUE, it prioritizes the test statistic more than the group size proportion.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

Either a vector of halting test functions (or function names) with the signature halting_test(condition, covariates, thresh) (for the meaning of the parameters see match_groups); or it may be a list of list(test = halting_test, cond = subset_of_conditions, cov = variable_selector, thresh) fields. All fields can be left out except test, and test need not be named if it is the first item in the list. The subset_of_conditions can be names of the conditions to match (a character vector or a factor). The variable_selector can be a logical vector with as many items as there will be columns in covariates (recommended), or a vector of integer covariate column indices. Each halting_test is then only applied to the specified subset of conditions and variables of the covariate matrix, with the specified threshold; when a value is not specified the defaults are used. Note that ordering the functions does not change the behavior, but can make the execution of the combined function faster, as the later ones are often evaluated only if the criteria for the earlier ones are met.

Assumes that at least a total of this many subjects will be preserved.

A factor vector containing condition labels.

Assumes that this number of cases are checked out per second, for estimating the time it takes to run the exhaustive search; default: 100.

If TRUE, prints partial calculations as well for the number of cases and estimated time when removing 1, 2, ... subjects.

A named integer vector, containing the maximum number of subjects that can be removed from each group. Specify 0 for groups if you want to preserve all of their subjects. If you do not specify a value for a group, it defaults to 2 less than the group size. Values outside the valid range of 0..(N-1) (where N is the number of subjects in the group) are corrected without a warning.

A particular set of group sizes that we know a matched solution for; min_preserved need not be specified if this one is.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used.

Once it is certain that the number of cases is definitely above this number, calculation stops. In this case, the returned number is guaranteed to be larger than max_cases, but it is not the exact number of exhaustive cases. Default is infimum, i.e. the exact number of cases is calculated.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

The name of the global parameter.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

The choice of search method, one of "random", You can get more information about each method on the help page for "search_<method_name>" (e.g. "search_exhaustive").

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

The maximum number of random replications to be performed. This is only used for the "random" method.

The minimum number of preserved subjects. It can be used to ensure that the search will not take forever to run, but instead fail when a solution is not found when preserving this number of subjects.

If TRUE, prints summary information on the input and the results, as well as progress information for the exhaustive search and random algorithms. Default: TRUE; can be changed using set_param("PRINT_INFO", FALSE).

A named integer vector, containing the maximum number of subjects that can be removed from each group. Specify 0 for groups if you want to preserve all of their subjects. If you do not specify a value for a group, it defaults to 2 less than the group size. Values outside the valid range of 0..(N-1) (where N is the number of subjects in the group) are corrected without a warning.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

The lookahead to use: a positive integer. It is used by the heuristic3 and heuristic4 algorithms, with a default of 2. The running time is O(N ^ lookahead), wheren N is the number of subjects.

If TRUE, returns all results found by method in a list. (A list is returned even if there is only one result.) If FALSE (the default), it returns the first result (a logical vector).

If TRUE, prefers higher test statistic more than the expected group size proportion; default is TRUE. Used by all algorithms except exhaustive, which always

The number of equivalent subjects that can be removed in each step. (The actual allowed number may be less depending on the p-value / theshold ratio.) This parameters is used by the heuristic3 and heuristic4 algorithms, with a default value of 1.

The percentage of remaining subjects that can be removed in each step. Used when max_removed_per_step > 1, with a default value of 0.5.

The p-value / threshold ratio at which it starts removing subjects one by one. Used when max_removed_per_step > 1, with a default value of 0.5.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

The maximum number of subjects that can be removed from each group. It must have a valid number for each group.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

The minimum number of preserved subjects. It can be used to ensure that the search will not take forever to run, but instead fail when a solution is not found when preserving this number of subjects.

If TRUE, prints summary information on the input and the results, as well as progress information for the exhaustive search and random algorithms. Default: TRUE; can be changed using set_param("PRINT_INFO", FALSE).

The names of arguments given to the search function.

Consumes extra parameters that are not used by the search algorithm at hand; this function gives a warning about the ones whose value is not NULL that their value is not used.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

The maximum number of subjects that can be removed from each group. It must have a valid number for each group.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

If TRUE, prefers higher test statistic more than the expected group size proportion; default is TRUE. Used by all algorithms except exhaustive, which always

If TRUE, prints summary information on the input and the results, as well as progress information for the exhaustive search and random algorithms. Default: TRUE; can be changed using set_param("PRINT_INFO", FALSE).

The names of arguments given to the search function.

Consumes extra parameters that are not used by the search algorithm at hand; this function gives a warning about the ones whose value is not NULL that their value is not used.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

The maximum number of subjects that can be removed from each group. It must have a valid number for each group.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

The minimum number of preserved subjects. It can be used to ensure that the search will not take forever to run, but instead fail when a solution is not found when preserving this number of subjects.

The lookahead to use: a positive integer. It is used by the heuristic3 and heuristic4 algorithms, with a default of 2. The running time is O(N ^ lookahead), wheren N is the number of subjects.

If TRUE, prefers higher test statistic more than the expected group size proportion; default is TRUE. Used by all algorithms except exhaustive, which always

If TRUE, prints summary information on the input and the results, as well as progress information for the exhaustive search and random algorithms. Default: TRUE; can be changed using set_param("PRINT_INFO", FALSE).

The number of equivalent subjects that can be removed in each step. (The actual allowed number may be less depending on the p-value / theshold ratio.) This parameters is used by the heuristic3 and heuristic4 algorithms, with a default value of 1.

The percentage of remaining subjects that can be removed in each step. Used when max_removed_per_step > 1, with a default value of 0.5.

The p-value / threshold ratio at which it starts removing subjects one by one. Used when max_removed_per_step > 1, with a default value of 0.5.

The names of arguments given to the search function.

Consumes extra parameters that are not used by the search algorithm at hand; this function gives a warning about the ones whose value is not NULL that their value is not used.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

The maximum number of subjects that can be removed from each group. It must have a valid number for each group.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

The minimum number of preserved subjects. It can be used to ensure that the search will not take forever to run, but instead fail when a solution is not found when preserving this number of subjects.

The lookahead to use: a positive integer. It is used by the heuristic3 and heuristic4 algorithms, with a default of 2. The running time is O(N ^ lookahead), wheren N is the number of subjects.

If TRUE, prefers higher test statistic more than the expected group size proportion; default is TRUE. Used by all algorithms except exhaustive, which always

If TRUE, prints summary information on the input and the results, as well as progress information for the exhaustive search and random algorithms. Default: TRUE; can be changed using set_param("PRINT_INFO", FALSE).

The number of equivalent subjects that can be removed in each step. (The actual allowed number may be less depending on the p-value / theshold ratio.) This parameters is used by the heuristic3 and heuristic4 algorithms, with a default value of 1.

The percentage of remaining subjects that can be removed in each step. Used when max_removed_per_step > 1, with a default value of 0.5.

The p-value / threshold ratio at which it starts removing subjects one by one. Used when max_removed_per_step > 1, with a default value of 0.5.

The names of arguments given to the search function.

Consumes extra parameters that are not used by the search algorithm at hand; this function gives a warning about the ones whose value is not NULL that their value is not used.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

A function to apply to 'covariates' (in matrix form) which is TRUE iff the conditions are matched. Signature: halting_test(condition, covariates, thresh). The following halting tests are part of this package: t_halt, U_halt, l_halt, ad_halt, ks_halt, wilks_halt, f_halt. You can create the intersection of two or more halting tests using create_halting_test.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

Either the desired proportions (percentage) of the sample for each condition as a named vector, or the names of the conditions for which we prefer to preserve the subjects, in decreasing order of preference. If not specified, the (full) sample proportions are used. This is preferred among configurations with the same taken into account by the other methods to some extent. For example, c(A = 0.4, B = 0.4, C = 0.2) means that we would like the number of subjects in groups A, B, and C to be around 40%, 40%, and 20% of the total number of subjects, respectively. Whereas c("A", "B", "C") means that if possible, we would like to keep all subjects in group A, and prefer keeping subjects in B, even if it results in losing more subjects from C.

The maximum number of subjects that can be removed from each group. It must have a valid number for each group.

NULL, or a function similar to halting_test, used to decide between cases for which halting_test yields equal values.

The maximum number of random replications to be performed. This is only used for the "random" method.

If TRUE, prefers higher test statistic more than the expected group size proportion; default is TRUE. Used by all algorithms except exhaustive, which always

If TRUE, prints summary information on the input and the results, as well as progress information for the exhaustive search and random algorithms. Default: TRUE; can be changed using set_param("PRINT_INFO", FALSE).

The names of arguments given to the search function.

Consumes extra parameters that are not used by the search algorithm at hand; this function gives a warning about the ones whose value is not NULL that their value is not used.

The name of the global parameter.

The new value of the global parameter.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.

A factor vector containing condition labels.

A columnwise matrix containing covariates to match the conditions on.

The return value of halting_test has to be greater than or equal to thresh for the matched groups.