getPredictedRRWpoints: This function simulates the RRW for a single set of...
In ccpluncw/ccpl_R_RRW: What the Package Does (One Line, Title Case)
getPredictedRRWpoints R Documentation
This function simulates the RRW for a single set of parameters across all values of overlap.
Description
This function simulates the RRW for a single set of parameters across all values of overlap and then attaches this simulation to the empirical RT and error data. This simulation's fit will be assessed by the optimization program (using assessRRWfit()) and then the parameters will be adjested.
Usage
getPredictedRRWpoints(
data,
RWkeepColumns,
mergeByDataColumns,
mergeByRWColumns,
dataRtCol,
RwSamplesCol,
dataOverlapCol,
b,
boundaryChangeSensitivity = 0.25,
Ter = 0,
startValue = 0,
noiseSD = 0,
decayBeta = 0,
decayAsymptote = 0.2,
valueChange = 0,
evaluationCriterion = 0,
bCols = NULL,
bcsCols = NULL,
TerCols = NULL,
sCols = NULL,
nSDCols = NULL,
dbCols = NULL,
daCols = NULL,
vcCols = NULL,
ecCols = NULL,
loops = 200
)
Arguments
data
This is a dataframe that must contain the following columns: overlap; RT (often a median); the proportion correct/incorrect; whether or not the row specifies a correct or incorrect trial. The dataset can also contains columns that effect code the influence of different parameters.
RWkeepColumns
A vector of strings that specify the columns of the RRW simulated data that should be kept when the simulated dataframe is merged with the empirical dataframe.
mergeByDataColumns
A vector of strings that specify the columns of the empirical data that should be used in the "by.x" option of the merge() function when the simulated dataframe is merged with the empirical dataframe.
mergeByRWColumns
A vector of strings that specify the columns of the RRW simulated data that should be used in the "by.y" option of the merge() function when the simulated dataframe is merged with the empirical dataframe.
dataRtCol
A string that identifies the name of the column in data that contains the RTs for the specific overlap/correct/condition combination. Default is "rt"
RwSamplesCol
A string that identifies the name of the column in the RRW simulations that contains the summary of the samples that you want to use as a simulation for RT. The possible columns are: "Q25" (the 25th quartile); "Q50" (the median); "mean" (the mean); "Q75" (the 75th quartile). The default is "Q50" because the median is more robust than the mean.
dataOverlapCol
A string that identifies the name of the column in data that contains the distributional overlaps for each row.
b
A vector of number(s) specifying the boundary distance from a 0 startpoint. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of b. The column names must be specified in the "bCols" argument. b is specific to the RRW simulation and has no default value
Ter
A vector of positive values that adjusts the time of encoding/response (non-decision time). If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect/dummy coded, because the values contained in this column will be multiplied by the value of Ter. All effect/dummy codes must >=0. The column names must be specified in the "TerCols" argument. When Ter = 0, a single, best fitting Ter will be calculated. Use this parameter when you have several conditions and you hypothesize that some conditions should have a larger Ter than others. For example, this may model longer encoding process when different numbers of items are on the screen for different conditions. The hypothetical fastest condition should have Ter = 0, which simply indicates that the RRW program will find the best fit Ter. The slower conditions will have larger Ter values and will be fit ralative to the fastest condition. The change in samples resulting from the Ter effect is a function of the size of the boundary (Ter * boundary). Default = 0, This is still experimental.
startValue
A vector of signed number(s) between -1 and 1 that indicates the position of the start point as a proportion of the boundary value. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of startValue. The column names must be specified in the "sCols" argument. startValue = 0 is the default and represents an unbiased start point.
noiseSD
A vector of positive number(s) representing the SD of the noise distribution. The noise distribution is N(0,nSD) and is added to the value of every step. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of noiseSD. The column names must be specified in the "nSDCols" argument. noiseSD = 0 is the default and represents no noise being added to each step.
decayBeta
A vector of signed number(s) representing the beta value in the Information Accrual Bias (IAB). If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of decayBeta. The column names must be specified in the "dbCols" argument. decayBeta = 0 is the default and represents no IAB.
decayAsymptote
A vector of positive number(s) representing the asymptote value in the Information Accrual Bias (IAB). If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of decayAsymptote. The column names must be specified in the "daCols" argument. decayAsymptote = 0.2 is the default.
valueChange
A vector of signed number(s) that indicates the change of overlap that one predicts. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of valueChange. The column names must be specified in the "vcCols" argument. The default is 0 because there is no valueChangeEffect.
evaluationCriterion
A vector of signed number(s) that indicates the change of the evaluation criterion (the position of the criterion in SDT) that one predicts. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of ec. The column names must be specified in the "ecCols" argument. The default is 0 because this is an unbiased evaluation criterion.
bCols
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of boundary (b). These columns should be effect coded, because the values contained in this column will be multiplied by the value of b.
bcsCols
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the boundaryChangeSensitivity (bcs). These columns must be effect coded, because the values contained in this column will be multiplied by the value of bcs. All effect/dummy codes must be >=0.
TerCols
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the non-decision time (Ter). These columns should be effect/dummy coded, because the values contained in this column will be multiplied by the value of Ter. All effect/dummy codes must be >=0.
sCols
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of startValue (startValue). These columns should be effect coded, because the values contained in this column will be multiplied by the value of startValue.
nSDCols
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of noiseSD (noiseSD). These columns should be effect coded, because the values contained in this column will be multiplied by the value of noiseSD.
dbCols
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of decayBeta (decayBeta). These columns should be effect coded, because the values contained in this column will be multiplied by the value of decayBeta.
daCols
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of decay asymptote (decayAsymptote). These columns should be effect coded, because the values contained in this column will be multiplied by the value of decayAsymptote.
vcCols
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of value change (valueChange). These columns should be effect coded, because the values contained in this column will be multiplied by the value of overlap.
ecCols
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of the evaluation criterion (ec). These columns should be effect coded, because the values contained in this column will be multiplied by the value of overlap.
loops
A number specifying the number of loops that will be run in the RRW simulation when it calculates the summary statistics for each number of samples for each boundary. Higher numbers produce more precise estimates, but also increase the time needed to converge on a solution. Default is 200.
bcs
A vector of positive values that specify the likelihood that the boundary will be reduced because there it is taking too many samples before reaching a threshold. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect/dummy coded, because the values contained in this column will be multiplied by the value of b. All effect/dummy codes must >=0. The column names must be specified in the "bcsCols" argument. Essentially, this models a decision-maker reducing the amount of information that they need before a response because it is taking to long. The boundary reduces when there is little change in the likelihood of a response after N number of steps. The routine makes this determination as a function of the size of the unaltered boundary (boundaryChangeSensitivity * boundary). Small values (e.g., 0.1), result in an impatient person (relatively early chnage in the boundary values). Large values (e.g., 0.9), result in a patient person (a boundary that is relatively resistent to change. A value of 0 is an exception: it will result in no boundary reduction under any circumstances. Impatience increases the likelihood of an error, so it increases the average time of error responses. Default is 0.25.
Value
A dataframe that contains the "data" plus the fitted values from the model ("Q25" (the 25th quartile); "Q50" (the median); "mean" (the mean); "Q75" (the 75th quartile); pCross (the fitted pHit from the model - probability of crossing each boundary); rtFit (the fitted rt values from the model))
Examples
getPredictedRRWpoints (data, RWkeepColumns = c("overlap", "Q50", "pCross", "correct"), mergeByDataColumns = c("overlap", "correct"), dataRtCol = "rt", RwSamplesCol = "Q50", dataOverlapCol= "overlap", b=14, s=0.1, loops = 400)
ccpluncw/ccpl_R_RRW documentation built on July 4, 2025, 3:24 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
| getPredictedRRWpoints | R Documentation |
This function simulates the RRW for a single set of parameters across all values of overlap.
Description
This function simulates the RRW for a single set of parameters across all values of overlap and then attaches this simulation to the empirical RT and error data. This simulation's fit will be assessed by the optimization program (using assessRRWfit()) and then the parameters will be adjested.
Usage
getPredictedRRWpoints(
data,
RWkeepColumns,
mergeByDataColumns,
mergeByRWColumns,
dataRtCol,
RwSamplesCol,
dataOverlapCol,
b,
boundaryChangeSensitivity = 0.25,
Ter = 0,
startValue = 0,
noiseSD = 0,
decayBeta = 0,
decayAsymptote = 0.2,
valueChange = 0,
evaluationCriterion = 0,
bCols = NULL,
bcsCols = NULL,
TerCols = NULL,
sCols = NULL,
nSDCols = NULL,
dbCols = NULL,
daCols = NULL,
vcCols = NULL,
ecCols = NULL,
loops = 200
)
Arguments
data |
This is a dataframe that must contain the following columns: overlap; RT (often a median); the proportion correct/incorrect; whether or not the row specifies a correct or incorrect trial. The dataset can also contains columns that effect code the influence of different parameters. |
RWkeepColumns |
A vector of strings that specify the columns of the RRW simulated data that should be kept when the simulated dataframe is merged with the empirical dataframe. |
mergeByDataColumns |
A vector of strings that specify the columns of the empirical data that should be used in the "by.x" option of the merge() function when the simulated dataframe is merged with the empirical dataframe. |
mergeByRWColumns |
A vector of strings that specify the columns of the RRW simulated data that should be used in the "by.y" option of the merge() function when the simulated dataframe is merged with the empirical dataframe. |
dataRtCol |
A string that identifies the name of the column in data that contains the RTs for the specific overlap/correct/condition combination. Default is "rt" |
RwSamplesCol |
A string that identifies the name of the column in the RRW simulations that contains the summary of the samples that you want to use as a simulation for RT. The possible columns are: "Q25" (the 25th quartile); "Q50" (the median); "mean" (the mean); "Q75" (the 75th quartile). The default is "Q50" because the median is more robust than the mean. |
dataOverlapCol |
A string that identifies the name of the column in data that contains the distributional overlaps for each row. |
b |
A vector of number(s) specifying the boundary distance from a 0 startpoint. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of b. The column names must be specified in the "bCols" argument. b is specific to the RRW simulation and has no default value |
Ter |
A vector of positive values that adjusts the time of encoding/response (non-decision time). If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect/dummy coded, because the values contained in this column will be multiplied by the value of Ter. All effect/dummy codes must >=0. The column names must be specified in the "TerCols" argument. When Ter = 0, a single, best fitting Ter will be calculated. Use this parameter when you have several conditions and you hypothesize that some conditions should have a larger Ter than others. For example, this may model longer encoding process when different numbers of items are on the screen for different conditions. The hypothetical fastest condition should have Ter = 0, which simply indicates that the RRW program will find the best fit Ter. The slower conditions will have larger Ter values and will be fit ralative to the fastest condition. The change in samples resulting from the Ter effect is a function of the size of the boundary (Ter * boundary). Default = 0, This is still experimental. |
startValue |
A vector of signed number(s) between -1 and 1 that indicates the position of the start point as a proportion of the boundary value. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of startValue. The column names must be specified in the "sCols" argument. startValue = 0 is the default and represents an unbiased start point. |
noiseSD |
A vector of positive number(s) representing the SD of the noise distribution. The noise distribution is N(0,nSD) and is added to the value of every step. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of noiseSD. The column names must be specified in the "nSDCols" argument. noiseSD = 0 is the default and represents no noise being added to each step. |
decayBeta |
A vector of signed number(s) representing the beta value in the Information Accrual Bias (IAB). If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of decayBeta. The column names must be specified in the "dbCols" argument. decayBeta = 0 is the default and represents no IAB. |
decayAsymptote |
A vector of positive number(s) representing the asymptote value in the Information Accrual Bias (IAB). If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of decayAsymptote. The column names must be specified in the "daCols" argument. decayAsymptote = 0.2 is the default. |
valueChange |
A vector of signed number(s) that indicates the change of overlap that one predicts. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of valueChange. The column names must be specified in the "vcCols" argument. The default is 0 because there is no valueChangeEffect. |
evaluationCriterion |
A vector of signed number(s) that indicates the change of the evaluation criterion (the position of the criterion in SDT) that one predicts. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect coded, because the values contained in this column will be multiplied by the value of ec. The column names must be specified in the "ecCols" argument. The default is 0 because this is an unbiased evaluation criterion. |
bCols |
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of boundary (b). These columns should be effect coded, because the values contained in this column will be multiplied by the value of b. |
bcsCols |
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the boundaryChangeSensitivity (bcs). These columns must be effect coded, because the values contained in this column will be multiplied by the value of bcs. All effect/dummy codes must be >=0. |
TerCols |
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the non-decision time (Ter). These columns should be effect/dummy coded, because the values contained in this column will be multiplied by the value of Ter. All effect/dummy codes must be >=0. |
sCols |
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of startValue (startValue). These columns should be effect coded, because the values contained in this column will be multiplied by the value of startValue. |
nSDCols |
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of noiseSD (noiseSD). These columns should be effect coded, because the values contained in this column will be multiplied by the value of noiseSD. |
dbCols |
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of decayBeta (decayBeta). These columns should be effect coded, because the values contained in this column will be multiplied by the value of decayBeta. |
daCols |
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of decay asymptote (decayAsymptote). These columns should be effect coded, because the values contained in this column will be multiplied by the value of decayAsymptote. |
vcCols |
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of value change (valueChange). These columns should be effect coded, because the values contained in this column will be multiplied by the value of overlap. |
ecCols |
A vector of strings that identifies the name(s) of the column in data that identifies the conditions that will affect the change of the evaluation criterion (ec). These columns should be effect coded, because the values contained in this column will be multiplied by the value of overlap. |
loops |
A number specifying the number of loops that will be run in the RRW simulation when it calculates the summary statistics for each number of samples for each boundary. Higher numbers produce more precise estimates, but also increase the time needed to converge on a solution. Default is 200. |
bcs |
A vector of positive values that specify the likelihood that the boundary will be reduced because there it is taking too many samples before reaching a threshold. If the number of values is greater than 1, then each value must have a corresponding effect coded column in the dataset. These columns should be effect/dummy coded, because the values contained in this column will be multiplied by the value of b. All effect/dummy codes must >=0. The column names must be specified in the "bcsCols" argument. Essentially, this models a decision-maker reducing the amount of information that they need before a response because it is taking to long. The boundary reduces when there is little change in the likelihood of a response after N number of steps. The routine makes this determination as a function of the size of the unaltered boundary (boundaryChangeSensitivity * boundary). Small values (e.g., 0.1), result in an impatient person (relatively early chnage in the boundary values). Large values (e.g., 0.9), result in a patient person (a boundary that is relatively resistent to change. A value of 0 is an exception: it will result in no boundary reduction under any circumstances. Impatience increases the likelihood of an error, so it increases the average time of error responses. Default is 0.25. |
Value
A dataframe that contains the "data" plus the fitted values from the model ("Q25" (the 25th quartile); "Q50" (the median); "mean" (the mean); "Q75" (the 75th quartile); pCross (the fitted pHit from the model - probability of crossing each boundary); rtFit (the fitted rt values from the model))
Examples
getPredictedRRWpoints (data, RWkeepColumns = c("overlap", "Q50", "pCross", "correct"), mergeByDataColumns = c("overlap", "correct"), dataRtCol = "rt", RwSamplesCol = "Q50", dataOverlapCol= "overlap", b=14, s=0.1, loops = 400)
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.