[R] Antwort: Re: Antwort: Re: Way to Plot Multiple Variables and Change Color
David Winsemius
dwinsemius at comcast.net
Mon Apr 10 22:21:46 CEST 2017
> On Apr 10, 2017, at 1:06 PM, David Winsemius <dwinsemius at comcast.net> wrote:
>
>
>> On Apr 10, 2017, at 7:45 AM, G.Maubach at weinwolf.de wrote:
>>
>> Hi Ulrik,
>>
>> many thanks for your reply. I had to take an unplanned break and was not
>> in the office during the last two weeks. Thus my late reply.
>>
>> I followed your advice and converted the variable in argument "fill" to
>> factor. Now the color change works:
>>
>> -- cut --
>>
>> d_result <- structure(list("variable" = c("Item 1 (ø = 3.3) ", "Item 1 (ø = 3.3) ",
>> "Item 1 (ø = 3.3) ", "Item 1 (ø = 3.3) ", "Item 1 (ø = 3.3) ",
>> "Item 1 (ø = 3.3) ", "Item 2 (ø = 3.8) ", "Item 2 (ø = 3.8) ",
>> "Item 2 (ø = 3.8) ", "Item 2 (ø = 3.8) ", "Item 2 (ø = 3.8) ",
>> "Item 2 (ø = 3.8) ", "Item 3 (ø = 3.4) ", "Item 3 (ø = 3.4) ",
>> "Item 3 (ø = 3.4) ", "Item 3 (ø = 3.4) ", "Item 3 (ø = 3.4) ",
>> "Item 3 (ø = 3.4) ", "Item 4 (ø = 3.4) ", "Item 4 (ø = 3.4) ",
>> "Item 4 (ø = 3.4) ", "Item 4 (ø = 3.4) ", "Item 4 (ø = 3.4) ",
>> "Item 4 (ø = 3.4) ", "Item 5 (ø = 3.5) ", "Item 5 (ø = 3.5) ",
>> "Item 5 (ø = 3.5) ", "Item 5 (ø = 3.5) ", "Item 5 (ø = 3.5) ",
>> "Item 5 (ø = 3.5) ", "Item 6 (ø = 3.5) ", "Item 6 (ø = 3.5) ",
>> "Item 6 (ø = 3.5) ", "Item 6 (ø = 3.5) ", "Item 6 (ø = 3.5) ",
>> "Item 6 (ø = 3.5) ", "Item 7 (ø = 3.4) ", "Item 7 (ø = 3.4) ",
>> "Item 7 (ø = 3.4) ", "Item 7 (ø = 3.4) ", "Item 7 (ø = 3.4) ",
>> "Item 7 (ø = 3.4) ", "Item 8 (ø = 3.3) ", "Item 8 (ø = 3.3) ",
>> "Item 8 (ø = 3.3) ", "Item 8 (ø = 3.3) ", "Item 8 (ø = 3.3) ",
>> "Item 8 (ø = 3.3) "), value =
>> structure(c(1L, 2L, 3L, 4L, 5L,
>> 6L, 1L, 2L, 3L, 4L, 5L, 6L, 1L, 2L, 3L, 4L, 5L, 6L, 1L, 2L, 3L,
>> 4L, 5L, 6L, 1L, 2L, 3L, 4L, 5L, 6L, 1L, 2L, 3L, 4L, 5L, 6L, 1L,
>> 2L, 3L, 4L, 5L, 6L, 1L, 2L, 3L, 4L, 5L, 6L), .Label = c("1 = very
>> satisfied",
>> "2", "3",
>> "4", "5", "6 = very dissatified"), class = "factor"),
>> n = c(14L, 20L, 24L, 14L, 16L, 14L, 9L, 15L,
>> 21L, 20L, 14L,
>> 23L, 19L, 17L, 16L, 14L, 16L, 20L, 22L,
>> 17L, 15L, 16L, 20L,
>> 12L, 19L, 15L, 16L, 15L, 18L, 19L, 18L,
>> 15L, 18L, 18L, 16L,
>> 17L, 17L, 20L, 17L, 17L, 14L, 16L, 16L,
>> 25L, 16L, 17L, 8L,
>> 20L)), .Names = c("variable", "value",
>> "n"), row.names =
>> c(NA,
>> -48L), vars = list("variable"), drop = TRUE,
>> indices =
>> list(0:5,
>> 6:11, 12:17, 18:23, 24:29, 30:35, 36:41,
>> 42:47),
>> group_sizes = c(6L,
>> 6L, 6L, 6L, 6L, 6L, 6L, 6L),
>> biggest_group_size = 6L,
>> labels = structure(list(
>> "variable" = structure(1:8, .Label = c("Item 1 (ø
>> = 3.3) ",
>> "Item 2 (ø =
>> 3.8) ", "Item 3 (ø = 3.4) ", "Item 4 (ø = 3.4) ",
>> "Item 5 (ø =
>> 3.5) ", "Item 6 (ø = 3.5) ", "Item 7 (ø = 3.4) ",
>> "Item 8 (ø =
>> 3.3) "), class = "factor")),
>> row.names = c(NA,
>> -8L), class = "data.frame", vars =
>> list("variable"),
>> drop = TRUE, .Names = "variable"),
>> class = c("grouped_df",
>> "tbl_df", "tbl", "data.frame"))
>>
>> ggplot(
>> d_result,
>> aes(x = variable, y = n, fill = rev(factor(value)))) +
>> geom_bar(
>> stat = "identity") +
>> coord_cartesian(ylim = c(0,100)) +
>> coord_flip() +
>> scale_y_continuous(name = "Percent") +
>> scale_fill_manual(
>> values = rev(
>> c(
>> "forestgreen", "limegreen",
>> "gold", "orange1",
>> "tomato3", "darkred"))) +
>> ggtitle(
>> paste(
>> "Question 8: Satisfaction?")) +
>> labs(fill = "Rating") +
>> scale_x_discrete(
>> name = element_blank()) +
>> # scale_color_manual(
>> # values = rev(
>> # c(
>> # "forestgreen", "limegreen",
>> # "gold", "orange1",
>> # "tomato3", "darkred"))) +
>> geom_text(
>> aes(label = n),
>> color = "white",
>> position = position_stack(vjust = 0.5)) +
>> theme_minimal() +
>> theme(
>> legend.position = "right")
>>
>> -- cut --
>>
>> I tried to change the order of the items on the y-axis, e.g. Item 8
>> should be last and Item 1 first.
>
> "First" and "last" apparently mean "top" and "bottom" to you. Since the $variable column is character, and ordering is typically done by setting levels of factors, try:
>
> d_result$variable <- factor(d_result$variable, levels=rev(unique(d_result$variable)))
>
>
> # changes ordering so the "Item 1"'s are at the top.
>
>
>> I tried to reverse the order of the items
>> within ggplot using rev() and relevel(). But neither of them worked. Is
>> there a way to do it?
>
> I don't think you can relevel a character column.
>>
>> I also tried to adjust the color palette for the legend, e.g. 1 = very
>> satisfied is green, 6 = very dissatified is red instead of vice versa as
>> it is now. The result should ensure the item naming for 1 = satisfied and
>> 6 = unsatifies cause this is the way it was asked in the questionnaire.
>>
>> Thus my question is:
>>
>> 1. How can I change the order of the sequence for the y-axis?
>>
>> 2. How can I adjust the color palette of the legend that it matches the
>> correct items?
>
> You probably could use relevel sinc `value` was a factor but I found it easier to simply repeat the relevelling code and change the target column name:
>
> d_result$value <- factor(d_result$value, levels=rev(unique(d_result$value)))
>
> I did find the appearance of the final result stange because there was irregular use of "\n" in the "variable" character values. that created more items than I think you wanted to appear.
I edited the code above to take out the inadvertent linefeeds (which got inserted by some part of the mail-processing chain) and then ran the ggplot call inside pdf() and print(...):
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Rplots.pdf
Type: application/pdf
Size: 5539 bytes
Desc: not available
URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20170410/346c3c9b/attachment.pdf>
-------------- next part --------------
HTH;
David.
>>
>> Can you give me a hint which functions I could use to do it?
>>
>> Kind regards
>>
>> Georg
>>
>>
>>
>>
>> Von: Ulrik Stervbo <ulrik.stervbo at gmail.com>
>> An: G.Maubach at weinwolf.de, "Richard M. Heiberger" <rmh at temple.edu>,
>> Kopie: r-help <r-help at r-project.org>
>> Datum: 28.03.2017 18:32
>> Betreff: Re: [R] Antwort: Re: Way to Plot Multiple Variables and
>> Change Color
>>
>>
>>
>> Hi Georg,
>>
>> you were on the right path - it is all about scale_fill*
>>
>> The 'problem' as you've discovered is that value is continuous, but
>> applying scale_fill_manual or others (except scale_fill_gradient) expects
>> discrete values.
>>
>> The solution is simply to set the fill with that by using factor():
>>
>> ggplot(
>> d_result,
>> aes(variable, y = n, fill = factor(value))) +
>> geom_bar(stat = "identity") +
>> scale_fill_manual(values = RColorBrewer::brewer.pal(4, "Blues"))
>> or:
>> ggplot(
>> d_result,
>> aes(variable, y = n, fill = factor(value))) +
>> geom_bar(stat = "identity") +
>> scale_fill_manual(values = c("red","blue", "green", "purple"))
>>
>> When using colorBrewer (which I highly recommend), I use scale_*_brewer
>> rather than setting the colour manually:
>>
>> ggplot(
>> d_result,
>> aes(variable, y = n, fill = factor(value))) +
>> geom_bar(stat = "identity") +
>> scale_fill_brewer(palette = "Blues ")
>>
>> Best,
>> Ulrik
>>
>>
>> On Tue, 28 Mar 2017 at 18:21 <G.Maubach at weinwolf.de> wrote:
>> Hi Richard,
>>
>> many thanks for your reply.
>>
>> Your solution is not exactly what I was looking for. I would like to know
>> how I can change the colors of the stacked bars in my plot and not use the
>> default values. How can this be done?
>>
>> Kind regards
>>
>> Georg
>>
>>
>>
>>
>> Von: "Richard M. Heiberger" <rmh at temple.edu>
>> An: G.Maubach at weinwolf.de,
>> Kopie: r-help <r-help at r-project.org>
>> Datum: 28.03.2017 17:40
>> Betreff: Re: [R] Way to Plot Multiple Variables and Change Color
>>
>>
>>
>> I think you are looking for the likert function in the HH package.
>>> From ?likert
>>
>>
>> Diverging stacked barcharts for Likert, semantic differential, rating
>> scale data, and population pyramids.
>>
>>
>> This will get you started. Much more fine control is available. See
>> the examples and demo.
>>
>> ## install.packages("HH") ## if not yet on your system.
>>
>> library(HH)
>>
>> AA <- dfr[,-9]
>>
>> labels <- sort(unique(as.vector(data.matrix(AA))))
>> result.template <- integer(length(labels))
>> names(result.template) <- labels
>>
>> BB <- apply(AA, 2, function(x, result=result.template) {
>> tx <- table(x)
>> result[names(tx)] <- tx
>> result
>> }
>> )
>>
>> BB
>>
>> likert(t(BB), ReferenceZero=0, horizontal=FALSE)
>>
>>
>> On Tue, Mar 28, 2017 at 6:05 AM, <G.Maubach at weinwolf.de> wrote:
>>> Hi All,
>>>
>>> in my current project I have to plot a whole bunch of related variables
>>> (item batteries, e.g. How do you rate ... a) Accelaration, b) Horse
>> Power,
>>> c) Color Palette, etc.) which are all rated on a scale from 1 .. 4.
>>>
>>> I need to present the results as stacked bar charts where the variables
>>> are columns and the percentages of the scales values (1 .. 4) are the
>>> chunks of the stacked bar for each variable. To do this I have
>> transformed
>>> my data from wide to long and calculated the percentage for each
>> variable
>>> and value. The code for this is as follows:
>>>
>>> -- cut --
>>>
>>> dfr <- structure(
>>> list(
>>> v07_01 = c(3, 1, 1, 4, 3, 4, 4, 1, 3, 2, 2, 3,
>>> 4, 4, 4, 1, 1, 3, 3, 4),
>>> v07_02 = c(1, 2, 1, 1, 2, 1, 4, 1, 1,
>>> 4, 4, 1, 4, 4, 1, 3, 2, 3, 3, 1),
>>> v07_03 = c(3, 2, 2, 1, 4, 1,
>>> 2, 3, 3, 1, 4, 2, 3, 1, 4, 1, 4, 2, 2, 3),
>>> v07_04 = c(3, 1, 1,
>>> 4, 2, 4, 4, 2, 2, 2, 4, 1, 2, 1, 3, 1, 2, 4, 1, 4),
>>> v07_05 = c(1,
>>> 2, 2, 2, 4, 4, 1, 1, 4, 4, 2, 1, 2, 1, 4, 1, 2, 4, 1, 4),
>>> v07_06 = c(1,
>>> 2, 1, 2, 1, 1, 3, 4, 3, 2, 2, 3, 3, 2, 4, 2, 3, 1, 4, 3),
>>> v07_07 = c(3,
>>> 2, 3, 3, 1, 1, 3, 3, 4, 4, 1, 3, 1, 3, 2, 4, 1, 2, 3, 4),
>>> v07_08 = c(3,
>>> 2, 1, 2, 2, 2, 3, 3, 4, 4, 1, 1, 1, 2, 3, 1, 4, 2, 2, 4),
>>> cased_id = structure(
>>> 1:20,
>>> .Label = c(
>>> "1",
>>> "2",
>>> "3",
>>> "4",
>>> "5",
>>> "6",
>>> "7",
>>> "8",
>>> "9",
>>> "10",
>>> "11",
>>> "12",
>>> "13",
>>> "14",
>>> "15",
>>> "16",
>>> "17",
>>> "18",
>>> "19",
>>> "20"
>>> ),
>>> class = "factor"
>>> )
>>> ),
>>> .Names = c(
>>> "v07_01",
>>> "v07_02",
>>> "v07_03",
>>> "v07_04",
>>> "v07_05",
>>> "v07_06",
>>> "v07_07",
>>> "v07_08",
>>> "cased_id"
>>> ),
>>> row.names = c(NA, -20L),
>>> class = c("tbl_df", "tbl",
>>> "data.frame")
>>> )
>>>
>>> mdf <- melt(df)
>>> d_result <- mdf %>%
>>> dplyr::group_by(variable) %>%
>>> count(value)
>>>
>>> ggplot(
>>> d_result,
>>> aes(variable, y = n, fill = value)) +
>>> geom_bar(stat = "identity") +
>>> coord_cartesian(ylim = c(0,100))
>>>
>>> -- cut --
>>>
>>> Is there an easier way of doing this, i. e. a way without need to
>>> transform the data?
>>>
>>> How can I change the colors for the data points 1 .. 4?
>>>
>>> I tried
>>>
>>> -- cut --
>>>
>>> d_result,
>>> aes(variable, y = n, fill = value)) +
>>> geom_bar(stat = "identity") +
>>> coord_cartesian(ylim = c(0,100)) +
>>> scale_fill_manual(values = RColorBrewer::brewer.pal(4, "Blues"))
>>>
>>> -- cut -
>>>
>>> but this does not work cause I am mixing continuous and descrete values.
>>>
>>> How can I change the colors for the bars?
>>>
>>> Kind regards
>>>
>>> Georg
>>>
>>> ______________________________________________
>>> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>> PLEASE do read the posting guide
>> http://www.R-project.org/posting-guide.html
>>> and provide commented, minimal, self-contained, reproducible code.
>>
>> ______________________________________________
>> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide
>> http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>
>> ______________________________________________
>> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>
> David Winsemius
> Alameda, CA, USA
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
David Winsemius
Alameda, CA, USA
More information about the R-help
mailing list