search

bnicenboim / eeguana

A package for manipulating EEG data in R.
https://bnicenboim.github.io/eeguana/
Other
21 stars 9 forks source link

default eeguana plot theme as a function? #68

Open stonekate opened 5 years ago

stonekate commented 5 years ago

Might be useful to have the default plot theme as a standalone plot function? At the moment you can manually make the theme via ggplot, e.g.:

# create fake dataset
data_1 <- eeg_lst(
  signal = signal_tbl(
    signal_matrix = as.matrix(
      data.frame(X = sin(1:30), Y = sin(1:30))
    ),
    ids = rep(c(1L, 2L, 3L), each = 10),
    sample_ids = sample_int(rep(seq(-4L, 5L), times = 3), sampling_rate = 500),
    dplyr::tibble(
      channel = c("X", "Y"), .reference = NA, theta = NA, phi = NA,
      radius = NA, .x = c(1, 1), .y = NA_real_, .z = NA_real_
    )
  ),
  events = dplyr::tribble(
    ~.id, ~type, ~description, ~.sample_0, ~.size, ~.channel,
    1L, "New Segment", NA_character_, -4L, 1L, NA,
    1L, "Bad", NA_character_, -2L, 3L, NA,
    1L, "Time 0", NA_character_, 1L, 1L, NA,
    1L, "Bad", NA_character_, 2L, 2L, "X",
    2L, "New Segment", NA_character_, -4L, 1L, NA,
    2L, "Time 0", NA_character_, 1L, 1L, NA,
    2L, "Bad", NA_character_, 2L, 1L, "Y",
    3L, "New Segment", NA_character_, -4L, 1L, NA,
    3L, "Time 0", NA_character_, 1L, 1L, NA,
    3L, "Bad", NA_character_, 2L, 1L, "Y"
  ),
  segments = dplyr::tibble(.id = c(1L, 2L, 3L),
                           recording = "recording1",
                           segment = c(1L, 2L, 3L),
                           condition = c("a", "b", "a"))
)

# just some different X and Y
data_2 <- mutate(data_1, recording = "recording2",
                 X = sin(X),
                 Y = sin(Y - .5),
                 condition = c("b", "a", "b"))

# bind it all together
data <- bind(data_1, data_2)

data %>%
  as_tibble() %>%
  ggplot2::ggplot(ggplot2::aes(time, amplitude)) + 
  ggplot2::geom_line(ggplot2::aes(group = .id, colour = condition), alpha = .5) +
  ggplot2::stat_summary(fun.y = "mean", geom = "line", ggplot2::aes(colour = condition), alpha = 1, size = 1) +
  ggplot2::facet_wrap(~ channel) +
  ggplot2::scale_colour_brewer(type = "qual", palette = "Dark2") +
  ggplot2::theme(legend.position = "bottom",
                 panel.background = ggplot2::element_blank(),
                 panel.spacing = ggplot2::unit(.01, "points"),
                 panel.grid = ggplot2::element_line(colour = "grey93"),
                 strip.background = ggplot2::element_blank())

But might be nice to just add something like (although theme_eeguana is currently something else):

data %>%
  as_tibble() %>%
  ggplot2::ggplot(ggplot2::aes(time, amplitude)) + 
  ggplot2::geom_line(ggplot2::aes(group = .id, colour = condition), alpha = .5) +
  ggplot2::stat_summary(fun.y = "mean", geom = "line", ggplot2::aes(colour = condition), alpha = 1, size = 1) +
  ggplot2::facet_wrap(~ channel) +
  theme_eeguana()
bnicenboim commented 5 years ago

you can do eeguana:::theme_eeguana

three : and no ()

On Sun, Mar 10, 2019, 12:26 auskate notifications@github.com wrote:

Might be useful to have the default plot theme as a standalone plot function? (at least for testing :) ), e.g.:

create fake dataset

data_1 <- eeg_lst( signal = signal_tbl( signal_matrix = as.matrix( data.frame(X = sin(1:30), Y = sin(1:30)) ), ids = rep(c(1L, 2L, 3L), each = 10), sample_ids = sample_int(rep(seq(-4L, 5L), times = 3), sampling_rate = 500), dplyr::tibble( channel = c("X", "Y"), .reference = NA, theta = NA, phi = NA, radius = NA, .x = c(1, 1), .y = NAreal, .z = NAreal ) ), events = dplyr::tribble( ~.id, ~type, ~description, ~.sample_0, ~.size, ~.channel, 1L, "New Segment", NAcharacter, -4L, 1L, NA, 1L, "Bad", NAcharacter, -2L, 3L, NA, 1L, "Time 0", NAcharacter, 1L, 1L, NA, 1L, "Bad", NAcharacter, 2L, 2L, "X", 2L, "New Segment", NAcharacter, -4L, 1L, NA, 2L, "Time 0", NAcharacter, 1L, 1L, NA, 2L, "Bad", NAcharacter, 2L, 1L, "Y", 3L, "New Segment", NAcharacter, -4L, 1L, NA, 3L, "Time 0", NAcharacter, 1L, 1L, NA, 3L, "Bad", NAcharacter, 2L, 1L, "Y" ), segments = dplyr::tibble(.id = c(1L, 2L, 3L), recording = "recording1", segment = c(1L, 2L, 3L), condition = c("a", "b", "a")) )

just some different X and Y

data_2 <- mutate(data_1, recording = "recording2", X = sin(X), Y = sin(Y - .5), condition = c("b", "a", "b"))

bind it all together

data <- bind(data_1, data_2)

data %>% as_tibble() %>% ggplot2::ggplot(ggplot2::aes(time, amplitude)) + ggplot2::geom_line(ggplot2::aes(group = .id, colour = condition), alpha = .5) + ggplot2::stat_summary(fun.y = "mean", geom = "line", aes(ggplot2::colour = condition), alpha = 1, size = 1) + ggplot2::facet_wrap(~ channel) + theme_eeguana()

or

eeguana::theme_eeguana()

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/bnicenboim/eeguana/issues/68, or mute the thread https://github.com/notifications/unsubscribe-auth/AFtIerHkst46Bzy3mL00iQti8UiGLDohks5vVOvtgaJpZM4bnNRH .

stonekate commented 5 years ago

Fancy! Would it be worth it to have it the ggplot way as well though? E.g. someone makes both default and tibble plots, and they want to make the tibble plots match the default plots. Not sure whether it's really necessary, was just a thought.

On Sun, Mar 10, 2019 at 8:48 PM Bruno Nicenboim notifications@github.com wrote:

you can do eeguana:::theme_eeguana

three : and no ()

On Sun, Mar 10, 2019, 12:26 auskate notifications@github.com wrote:

Might be useful to have the default plot theme as a standalone plot function? (at least for testing :) ), e.g.:

create fake dataset

data_1 <- eeg_lst( signal = signal_tbl( signal_matrix = as.matrix( data.frame(X = sin(1:30), Y = sin(1:30)) ), ids = rep(c(1L, 2L, 3L), each = 10), sample_ids = sample_int(rep(seq(-4L, 5L), times = 3), sampling_rate = 500), dplyr::tibble( channel = c("X", "Y"), .reference = NA, theta = NA, phi = NA, radius = NA, .x = c(1, 1), .y = NAreal, .z = NAreal ) ), events = dplyr::tribble( ~.id, ~type, ~description, ~.sample_0, ~.size, ~.channel, 1L, "New Segment", NAcharacter, -4L, 1L, NA, 1L, "Bad", NAcharacter, -2L, 3L, NA, 1L, "Time 0", NAcharacter, 1L, 1L, NA, 1L, "Bad", NAcharacter, 2L, 2L, "X", 2L, "New Segment", NAcharacter, -4L, 1L, NA, 2L, "Time 0", NAcharacter, 1L, 1L, NA, 2L, "Bad", NAcharacter, 2L, 1L, "Y", 3L, "New Segment", NAcharacter, -4L, 1L, NA, 3L, "Time 0", NAcharacter, 1L, 1L, NA, 3L, "Bad", NAcharacter, 2L, 1L, "Y" ), segments = dplyr::tibble(.id = c(1L, 2L, 3L), recording = "recording1", segment = c(1L, 2L, 3L), condition = c("a", "b", "a")) )

just some different X and Y

data_2 <- mutate(data_1, recording = "recording2", X = sin(X), Y = sin(Y - .5), condition = c("b", "a", "b"))

bind it all together

data <- bind(data_1, data_2)

data %>% as_tibble() %>% ggplot2::ggplot(ggplot2::aes(time, amplitude)) + ggplot2::geom_line(ggplot2::aes(group = .id, colour = condition), alpha = .5) + ggplot2::stat_summary(fun.y = "mean", geom = "line", aes(ggplot2::colour = condition), alpha = 1, size = 1) + ggplot2::facet_wrap(~ channel) + theme_eeguana()

or

eeguana::theme_eeguana()

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/bnicenboim/eeguana/issues/68, or mute the thread < https://github.com/notifications/unsubscribe-auth/AFtIerHkst46Bzy3mL00iQti8UiGLDohks5vVOvtgaJpZM4bnNRH

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/bnicenboim/eeguana/issues/68#issuecomment-471337739, or mute the thread https://github.com/notifications/unsubscribe-auth/AMJVQR-QGIXM0IZq6UXnuGh3hSS70D3Hks5vVWGOgaJpZM4bnNRH .

-- Kate Stone PhD candidate Vasishth Lab | Department of Linguistics Potsdam University, 14467 Potsdam, Germany https://auskate.github.io