R/pigeon_weber.R
In NourAl-Zaghloul/PigeonAnalyzeR:
Defines functions
pigeon_weber
# Calculates Weber scores (translated to r from Price et al. 2004 formulae)
pigeon_weber <- function(x, ...){
# Requires pracma for as close as possible exact emulation, but it shouldn't matter
# TODO: check if pracma is installed & ask to install if not
# TODO: Also for magrittr & dplyr
# TODO: Check if ratio is Large/Small or Small/Large and convert if necessary
# TODO: Make sure package functions are called explicitly
#### Sets the default column #s for participant, ratio, correct, and trialtype ----
# Requires the participant, ratio, and correct, but not trialtype
# TODO: Re-integrate trialtype into function (removed for simplicity)
# !. Currently users can separate the trialtypes into parts of a list and then *lapply*
if(!exists("weber_participant")){
weber_participant = 1
}
if(!exists("weber_ratio")){
weber_ratio = 2
}
if(!exists("weber_correct")){
weber_correct = 3
}
# if(is.null(weber_trialtype) & length(x) >= 4){
# weber_trialtype = 4
# }
#### Creates our summarised dataframe----
colnames(x)[c(weber_participant, weber_ratio, weber_correct)] <- c("part","ratio","correct")
x <- x %>%
dplyr::group_by(part, ratio) %>%
dplyr::summarise(
accuracy = mean(correct),
freq = n()
)
x_participants <- unique(x$part)
# TODO: ask if they want to set a anchorpoint ratios (i.e 1:1 ratio = .5 accuracy)
# !. does it by default as is
x_anchor <- data.frame(
part = x_participants,
ratio = 1,
freq = 1,
accuracy = .5)
x <- dplyr::bind_rows(x, x_anchor)
#### Creating necessary functions ----
# remaking the erfc function to replicate the matlab one
# shouldn't make a difference b/w this and pracma::erfc but just in case...
erfc2 <- function(y){
OUT <- 2*pnorm(y*sqrt(2), lower.tail = FALSE)
return(OUT)
}
# Creating the ANSmodel function
ANSModel <- function(weber){
OUT <- rep(0, length(ratio))
for(j in seq(length(ratio))){
OUT[j] <-
1 - (0.5 * erfc2(abs(ratio[j]-1)/( sqrt(2)*weber*sqrt(ratio[j]^2 + 1))))
}
return(OUT)
}
# Creating the lsqWeber function
lsqWeber <- function(weber){
modelY <- ANSModel(weber)
z <- subAcc - modelY
return(z)
}
#### w for each participant ----
# creates weber fraction output & cor with model ("Results" + "pearR" in matlab)
weber_w <- data.frame(
Participant = x_participants,
Results = rep(0, length(x_participants)),
Correlation = rep(0,length(x_participants))
)
# creates an overview of the processed data ("AccArray" in matlab)
weber_overview <- data.frame(
Participant = character(),
Ratio = double(),
SubjectAccuracy = double(),
W = double(),
ModelAccuracy = double(),
ssq = double(),
Residual = double()
)
# Getting the w score for each participant
# TODO: Change to an apply version if possible?
for(i in seq(length(x_participants))){
# Taking only a single participant's data
weber_currentpart <- x[x$part == x_participants[i],]
# creating vectors for Ratio & Accuracy
# TODO: Clean up what they wrote not have so much redundancy (except where helpful)
ratio <- weber_currentpart$ratio
subAcc <- weber_currentpart$accuracy
# perfoming least squares non-linear regression
#. tolx change to match matlab default (though doesn't matter for this granularity)
x_results <- lsqnonlin(lsqWeber,.25, options = list(tolx = 1e-06))
# creating what we would predict if model was 100% accurate
x_modelAccuracy <- ANSModel(x_results[[1]])
# creates dataframe which lists the participant, their w, and correlation between
#. their w and the model's prediction
weber_w[i,2] <- x_results[[1]]
weber_w[i,3] <- cor(subAcc, x_modelAccuracy)
# Larger dataframe showing the nuances of their results
#. Missing Residuals (function doesn't give them)
# TODO: Figure out how to get residuals out of lsqnonlin()
x_overviewtemp <- data.frame(
Participant = rep(x_participants[i], length(ratio)),
Ratio = ratio,
SubjectAccuracy = subAcc,
W = rep(x_results[[1]],length(ratio)),
ModelAccuracy = x_modelAccuracy,
ssq = x_results[[2]],
Residual = rep(NA_integer_,length(ratio))
)
# Attaches new data to the df with every participant
weber_overview <- bind_rows(weber_overview, x_overviewtemp)
}
return(list(weber_w, weber_overview))
}
#### Sample Data----
# sampledata <- data.frame(stringsAsFactors=FALSE,
# part = c("Gramps", "Gramps", "Gramps", "Gramps", "Gramps", "Gramps",
# "Gramps", "Gramps", "Gramps", "Gramps", "Gramps", "Gramps",
# "Negligble", "Negligble", "Negligble", "Negligble", "Negligble",
# "Negligble", "Negligble", "Negligble", "Negligble", "Negligble",
# "Negligble", "Negligble", "Whosit", "Whosit", "Whosit",
# "Whosit", "Whosit", "Whosit", "Whosit", "Whosit", "Whosit", "Whosit",
# "Whosit", "Whosit", "WhatsApp", "WhatsApp", "WhatsApp",
# "WhatsApp", "WhatsApp", "WhatsApp", "WhatsApp", "WhatsApp", "WhatsApp",
# "WhatsApp", "WhatsApp", "WhatsApp", "Yalla", "Yalla", "Yalla",
# "Yalla", "Yalla", "Yalla", "Yalla", "Yalla", "Yalla", "Yalla",
# "Yalla", "Yalla"),
# ratio = c(0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8, 0.85, 0.9,
# 0.95, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8, 0.85,
# 0.9, 0.95, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8, 0.85,
# 0.9, 0.95, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8,
# 0.85, 0.9, 0.95, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8,
# 0.85, 0.9, 0.95),
# correct = c(1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1,
# 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1,
# 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1)
# )
#### Original Code ----
# #### Introductions
# # This script does the actual analysis of the clean data:
# # Weber scores (translated to r from Price et al. 2004 formulae)
# # As closely emulating
#
# # packages:
# library(tidyverse) # Shouldn't need to load because source(0_...) would load it
# library(pracma) #for the equivalent to the matlab erfc and lsqnonlin functions
#
# # This script assumes that you'll receive long data of a 2AFC study (one line per answer)
# # "Ratio","participant", and "correct" need to be column names
# # Ratio = number between 0 and 1 that is the ratio of the numbers displayed
# # participant = participantID
# # correct = 1 or 0, if they got the trial right or not
#
# # Data = DotData & NaturalData
#
# #### Creating Data for weber
#
# # Getting all unique participant IDs
# participants <- unique(c(DotData$participant, NaturalData$participant))
#
# # Collapses data by ratio and participant
# #. Calculate mean accuracy at a ratio & frequency ratio occured
# D_ratioXaccuracy_participant <- DotData %>%
# group_by(Ratio, participant) %>%
# summarise(accuracy = mean(correct),
# freq = n()) %>%
# mutate(arrayType = "Dot")
#
# N_ratioXaccuracy_participant <- NaturalData %>%
# group_by(Ratio, participant) %>%
# summarise(accuracy = mean(correct),
# freq = n()) %>%
# mutate(arrayType = "Natural")
#
# # adding in anchorpoints for each participant
# D_anchor <- tibble(
# participant = participants,
# Ratio = 1,
# freq = 1,
# accuracy = .5,
# arrayType = 'Dot')
# D_ratioXaccuracy_participant <- bind_rows(D_ratioXaccuracy_participant, D_anchor)
#
# N_anchor <- tibble(
# participant = participants,
# Ratio = 1,
# freq = 1,
# accuracy = .5,
# arrayType = 'Natural')
# N_ratioXaccuracy_participant <- bind_rows(N_ratioXaccuracy_participant, N_anchor)
#
# #### Creating necessary functions
#
# # remaking the erfc function to replicate the matlab one
# # shouldn't make a difference b/w this and pracma::erfc but just in case...
# erfc2 <- function(x){
# OUT <- 2*pnorm(x*sqrt(2), lower.tail = FALSE)
# return(OUT)
# }
#
# # Creating the ANSmodel function
# ANSModel <- function(weber){
# OUT <- rep(0, length(ratio))
# for(j in seq(length(ratio))){
# OUT[j] <-
# 1 - (0.5 * erfc2(abs(ratio[j]-1)/( sqrt(2)*weber*sqrt(ratio[j]^2 + 1))))
# }
# return(OUT)
# }
#
# # Creating the lsqWeber function
# lsqWeber <- function(weber){
# modelY <- ANSModel(weber)
# y <- subAcc - modelY
# return(y)
# }
#
# #### w for each participant
#
# # creates weber fraction output & cor with model ("Results" + "pearR" in matlab)
# D_w <- tibble(
# participants = participants,
# results = rep(0, length(participants)),
# correlation = rep(0,length(participants))
# )
# N_w <- tibble(
# participants = participants,
# results = rep(0, length(participants)),
# correlation = rep(0,length(participants))
# )
#
# # creates an overview of the processed data ("AccArray" in matlab)
# D_overview <- tibble(
# Participant = character(),
# Ratio = double(),
# SubjectAccuracy = double(),
# W = double(),
# ModelAccuracy = double(),
# ssq = double(),
# Residual = double()
# )
# N_overview <- tibble(
# Participant = character(),
# Ratio = double(),
# SubjectAccuracy = double(),
# W = double(),
# ModelAccuracy = double(),
# ssq = double(),
# Residual = double()
# )
#
#
# # Getting the w score for each participant
# for(i in 1:length(participants)){
#
# # Taking only a single participant's data
# D_participant <- D_ratioXaccuracy_participant[
# D_ratioXaccuracy_participant$participant == participants[i],]
# N_participant <- N_ratioXaccuracy_participant[
# N_ratioXaccuracy_participant$participant == participants[i],]
#
# # creating vectors for Ratio & Accuracy
# ratio <- D_participant$Ratio
# subAcc <- D_participant$accuracy
#
# # perfoming least squares non-linear regression
# #. tolx change to match matlab default (though doesn't matter for this granularity)
# D_results <- lsqnonlin(lsqWeber,.25, options = list(tolx = 1e-06))
# # creating what we would predict if model was 100% accurate
# D_modelAccuracy <- ANSModel(D_results[[1]])
# # creates dataframe which lists the participant, their w, and correlation between
# #. their w and the model's prediction
# D_w[i,2] <- D_results[[1]]
# D_w[i,3] <- cor(subAcc, D_modelAccuracy)
# # Larger dataframe showing the nuances of their results
# #. Missing Residuals (function doesn't give them)
# # TODO: Figure out how to get residuals out of lsqnonlin()
# D_overviewtemp <- tibble(
# Participant = rep(participants[i], length(ratio)),
# Ratio = ratio,
# SubjectAccuracy = subAcc,
# W = rep(D_results[[1]],length(ratio)),
# ModelAccuracy = D_modelAccuracy,
# ssq = D_results[[2]],
# Residual = rep(NA_integer_,length(ratio))
# )
# # Attaches new data to the df with every participant
# D_overview <- bind_rows(D_overview, D_overviewtemp)
#
#
# # repeats the same thing as above, but for the other test type
# ratio <- N_participant$Ratio
# subAcc <- N_participant$accuracy
# N_results <- lsqnonlin(lsqWeber,.1)
# N_modelAccuracy <- ANSModel(N_results[[1]])
# N_w[i,2] <- N_results[[1]]
# N_w[i,3] <- cor(subAcc, N_modelAccuracy)
# N_overviewtemp <- tibble(
# Participant = rep(participants[i], length(ratio)),
# Ratio = ratio,
# SubjectAccuracy = subAcc,
# W = rep(N_results[[1]],length(ratio)),
# ModelAccuracy = N_modelAccuracy,
# ssq = N_results[[2]],
# Residual = rep(NA_integer_,length(ratio))
# )
# N_overview <- bind_rows(N_overview, N_overviewtemp)
# }
# DONE! (:
NourAl-Zaghloul/PigeonAnalyzeR documentation built on Sept. 12, 2019, 5:46 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.
Defines functions pigeon_weber
# Calculates Weber scores (translated to r from Price et al. 2004 formulae)
pigeon_weber <- function(x, ...){
# Requires pracma for as close as possible exact emulation, but it shouldn't matter
# TODO: check if pracma is installed & ask to install if not
# TODO: Also for magrittr & dplyr
# TODO: Check if ratio is Large/Small or Small/Large and convert if necessary
# TODO: Make sure package functions are called explicitly
#### Sets the default column #s for participant, ratio, correct, and trialtype ----
# Requires the participant, ratio, and correct, but not trialtype
# TODO: Re-integrate trialtype into function (removed for simplicity)
# !. Currently users can separate the trialtypes into parts of a list and then *lapply*
if(!exists("weber_participant")){
weber_participant = 1
}
if(!exists("weber_ratio")){
weber_ratio = 2
}
if(!exists("weber_correct")){
weber_correct = 3
}
# if(is.null(weber_trialtype) & length(x) >= 4){
# weber_trialtype = 4
# }
#### Creates our summarised dataframe----
colnames(x)[c(weber_participant, weber_ratio, weber_correct)] <- c("part","ratio","correct")
x <- x %>%
dplyr::group_by(part, ratio) %>%
dplyr::summarise(
accuracy = mean(correct),
freq = n()
)
x_participants <- unique(x$part)
# TODO: ask if they want to set a anchorpoint ratios (i.e 1:1 ratio = .5 accuracy)
# !. does it by default as is
x_anchor <- data.frame(
part = x_participants,
ratio = 1,
freq = 1,
accuracy = .5)
x <- dplyr::bind_rows(x, x_anchor)
#### Creating necessary functions ----
# remaking the erfc function to replicate the matlab one
# shouldn't make a difference b/w this and pracma::erfc but just in case...
erfc2 <- function(y){
OUT <- 2*pnorm(y*sqrt(2), lower.tail = FALSE)
return(OUT)
}
# Creating the ANSmodel function
ANSModel <- function(weber){
OUT <- rep(0, length(ratio))
for(j in seq(length(ratio))){
OUT[j] <-
1 - (0.5 * erfc2(abs(ratio[j]-1)/( sqrt(2)*weber*sqrt(ratio[j]^2 + 1))))
}
return(OUT)
}
# Creating the lsqWeber function
lsqWeber <- function(weber){
modelY <- ANSModel(weber)
z <- subAcc - modelY
return(z)
}
#### w for each participant ----
# creates weber fraction output & cor with model ("Results" + "pearR" in matlab)
weber_w <- data.frame(
Participant = x_participants,
Results = rep(0, length(x_participants)),
Correlation = rep(0,length(x_participants))
)
# creates an overview of the processed data ("AccArray" in matlab)
weber_overview <- data.frame(
Participant = character(),
Ratio = double(),
SubjectAccuracy = double(),
W = double(),
ModelAccuracy = double(),
ssq = double(),
Residual = double()
)
# Getting the w score for each participant
# TODO: Change to an apply version if possible?
for(i in seq(length(x_participants))){
# Taking only a single participant's data
weber_currentpart <- x[x$part == x_participants[i],]
# creating vectors for Ratio & Accuracy
# TODO: Clean up what they wrote not have so much redundancy (except where helpful)
ratio <- weber_currentpart$ratio
subAcc <- weber_currentpart$accuracy
# perfoming least squares non-linear regression
#. tolx change to match matlab default (though doesn't matter for this granularity)
x_results <- lsqnonlin(lsqWeber,.25, options = list(tolx = 1e-06))
# creating what we would predict if model was 100% accurate
x_modelAccuracy <- ANSModel(x_results[[1]])
# creates dataframe which lists the participant, their w, and correlation between
#. their w and the model's prediction
weber_w[i,2] <- x_results[[1]]
weber_w[i,3] <- cor(subAcc, x_modelAccuracy)
# Larger dataframe showing the nuances of their results
#. Missing Residuals (function doesn't give them)
# TODO: Figure out how to get residuals out of lsqnonlin()
x_overviewtemp <- data.frame(
Participant = rep(x_participants[i], length(ratio)),
Ratio = ratio,
SubjectAccuracy = subAcc,
W = rep(x_results[[1]],length(ratio)),
ModelAccuracy = x_modelAccuracy,
ssq = x_results[[2]],
Residual = rep(NA_integer_,length(ratio))
)
# Attaches new data to the df with every participant
weber_overview <- bind_rows(weber_overview, x_overviewtemp)
}
return(list(weber_w, weber_overview))
}
#### Sample Data----
# sampledata <- data.frame(stringsAsFactors=FALSE,
# part = c("Gramps", "Gramps", "Gramps", "Gramps", "Gramps", "Gramps",
# "Gramps", "Gramps", "Gramps", "Gramps", "Gramps", "Gramps",
# "Negligble", "Negligble", "Negligble", "Negligble", "Negligble",
# "Negligble", "Negligble", "Negligble", "Negligble", "Negligble",
# "Negligble", "Negligble", "Whosit", "Whosit", "Whosit",
# "Whosit", "Whosit", "Whosit", "Whosit", "Whosit", "Whosit", "Whosit",
# "Whosit", "Whosit", "WhatsApp", "WhatsApp", "WhatsApp",
# "WhatsApp", "WhatsApp", "WhatsApp", "WhatsApp", "WhatsApp", "WhatsApp",
# "WhatsApp", "WhatsApp", "WhatsApp", "Yalla", "Yalla", "Yalla",
# "Yalla", "Yalla", "Yalla", "Yalla", "Yalla", "Yalla", "Yalla",
# "Yalla", "Yalla"),
# ratio = c(0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8, 0.85, 0.9,
# 0.95, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8, 0.85,
# 0.9, 0.95, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8, 0.85,
# 0.9, 0.95, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8,
# 0.85, 0.9, 0.95, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8,
# 0.85, 0.9, 0.95),
# correct = c(1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1,
# 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1,
# 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1)
# )
#### Original Code ----
# #### Introductions
# # This script does the actual analysis of the clean data:
# # Weber scores (translated to r from Price et al. 2004 formulae)
# # As closely emulating
#
# # packages:
# library(tidyverse) # Shouldn't need to load because source(0_...) would load it
# library(pracma) #for the equivalent to the matlab erfc and lsqnonlin functions
#
# # This script assumes that you'll receive long data of a 2AFC study (one line per answer)
# # "Ratio","participant", and "correct" need to be column names
# # Ratio = number between 0 and 1 that is the ratio of the numbers displayed
# # participant = participantID
# # correct = 1 or 0, if they got the trial right or not
#
# # Data = DotData & NaturalData
#
# #### Creating Data for weber
#
# # Getting all unique participant IDs
# participants <- unique(c(DotData$participant, NaturalData$participant))
#
# # Collapses data by ratio and participant
# #. Calculate mean accuracy at a ratio & frequency ratio occured
# D_ratioXaccuracy_participant <- DotData %>%
# group_by(Ratio, participant) %>%
# summarise(accuracy = mean(correct),
# freq = n()) %>%
# mutate(arrayType = "Dot")
#
# N_ratioXaccuracy_participant <- NaturalData %>%
# group_by(Ratio, participant) %>%
# summarise(accuracy = mean(correct),
# freq = n()) %>%
# mutate(arrayType = "Natural")
#
# # adding in anchorpoints for each participant
# D_anchor <- tibble(
# participant = participants,
# Ratio = 1,
# freq = 1,
# accuracy = .5,
# arrayType = 'Dot')
# D_ratioXaccuracy_participant <- bind_rows(D_ratioXaccuracy_participant, D_anchor)
#
# N_anchor <- tibble(
# participant = participants,
# Ratio = 1,
# freq = 1,
# accuracy = .5,
# arrayType = 'Natural')
# N_ratioXaccuracy_participant <- bind_rows(N_ratioXaccuracy_participant, N_anchor)
#
# #### Creating necessary functions
#
# # remaking the erfc function to replicate the matlab one
# # shouldn't make a difference b/w this and pracma::erfc but just in case...
# erfc2 <- function(x){
# OUT <- 2*pnorm(x*sqrt(2), lower.tail = FALSE)
# return(OUT)
# }
#
# # Creating the ANSmodel function
# ANSModel <- function(weber){
# OUT <- rep(0, length(ratio))
# for(j in seq(length(ratio))){
# OUT[j] <-
# 1 - (0.5 * erfc2(abs(ratio[j]-1)/( sqrt(2)*weber*sqrt(ratio[j]^2 + 1))))
# }
# return(OUT)
# }
#
# # Creating the lsqWeber function
# lsqWeber <- function(weber){
# modelY <- ANSModel(weber)
# y <- subAcc - modelY
# return(y)
# }
#
# #### w for each participant
#
# # creates weber fraction output & cor with model ("Results" + "pearR" in matlab)
# D_w <- tibble(
# participants = participants,
# results = rep(0, length(participants)),
# correlation = rep(0,length(participants))
# )
# N_w <- tibble(
# participants = participants,
# results = rep(0, length(participants)),
# correlation = rep(0,length(participants))
# )
#
# # creates an overview of the processed data ("AccArray" in matlab)
# D_overview <- tibble(
# Participant = character(),
# Ratio = double(),
# SubjectAccuracy = double(),
# W = double(),
# ModelAccuracy = double(),
# ssq = double(),
# Residual = double()
# )
# N_overview <- tibble(
# Participant = character(),
# Ratio = double(),
# SubjectAccuracy = double(),
# W = double(),
# ModelAccuracy = double(),
# ssq = double(),
# Residual = double()
# )
#
#
# # Getting the w score for each participant
# for(i in 1:length(participants)){
#
# # Taking only a single participant's data
# D_participant <- D_ratioXaccuracy_participant[
# D_ratioXaccuracy_participant$participant == participants[i],]
# N_participant <- N_ratioXaccuracy_participant[
# N_ratioXaccuracy_participant$participant == participants[i],]
#
# # creating vectors for Ratio & Accuracy
# ratio <- D_participant$Ratio
# subAcc <- D_participant$accuracy
#
# # perfoming least squares non-linear regression
# #. tolx change to match matlab default (though doesn't matter for this granularity)
# D_results <- lsqnonlin(lsqWeber,.25, options = list(tolx = 1e-06))
# # creating what we would predict if model was 100% accurate
# D_modelAccuracy <- ANSModel(D_results[[1]])
# # creates dataframe which lists the participant, their w, and correlation between
# #. their w and the model's prediction
# D_w[i,2] <- D_results[[1]]
# D_w[i,3] <- cor(subAcc, D_modelAccuracy)
# # Larger dataframe showing the nuances of their results
# #. Missing Residuals (function doesn't give them)
# # TODO: Figure out how to get residuals out of lsqnonlin()
# D_overviewtemp <- tibble(
# Participant = rep(participants[i], length(ratio)),
# Ratio = ratio,
# SubjectAccuracy = subAcc,
# W = rep(D_results[[1]],length(ratio)),
# ModelAccuracy = D_modelAccuracy,
# ssq = D_results[[2]],
# Residual = rep(NA_integer_,length(ratio))
# )
# # Attaches new data to the df with every participant
# D_overview <- bind_rows(D_overview, D_overviewtemp)
#
#
# # repeats the same thing as above, but for the other test type
# ratio <- N_participant$Ratio
# subAcc <- N_participant$accuracy
# N_results <- lsqnonlin(lsqWeber,.1)
# N_modelAccuracy <- ANSModel(N_results[[1]])
# N_w[i,2] <- N_results[[1]]
# N_w[i,3] <- cor(subAcc, N_modelAccuracy)
# N_overviewtemp <- tibble(
# Participant = rep(participants[i], length(ratio)),
# Ratio = ratio,
# SubjectAccuracy = subAcc,
# W = rep(N_results[[1]],length(ratio)),
# ModelAccuracy = N_modelAccuracy,
# ssq = N_results[[2]],
# Residual = rep(NA_integer_,length(ratio))
# )
# N_overview <- bind_rows(N_overview, N_overviewtemp)
# }
# DONE! (:
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.