Hi Dennis,
Thank you for your constructive advice on plotting CIs in point ranges. I
combined the advices from you, Jim, Kaori, and bosie (from IRC) and my own
research and produced the following figure:
WTP_labels <- c("Wild cod", "Farmed cod", "Farmed salmon", "Wild monk",
"Farmed pangasius", "MSC label", "AB label")
plot_data <- data.frame(
xmin = c(1, 1.5, 3, 3.5, 5, 5.5, 7, 7.5, 9, 9.5, 11, 11.5, 13, 13.5)
, ymin = c(16.7026, 14.9968, 16.0630, 17.7510, -5.01694, -.44146,
1.45884
, 17.20, 16.32, 15.86, 18.12, -8.86, -0.03, 0.95)
, ymax = c(21.1602, 18.7613, 19.1367, 23.6730, 2.26564, 3.08630,
3.39865
, 20.12, 18.60, 18.14, 22.29, 3.03, 3.57, 2.81)
, ymean = c(18.93, 16.88, 17.60, 20.71, -1.38, 1.32, 2.43
, 18.66, 17.46, 17.00, 20.20, -2.92, 1.77, 1.88)
, model = rep(c("MMNL", "GMNL"), each = 7)
, id = factor(WTP_labels, levels = WTP_labels)
)
plot_data$"xmax" <- plot_data$"xmin" + 1
WTP_plot_pointrange <- ggplot(plot_data, aes(x = id, y = ymean, color =
model)) +
theme_bw() +
# geom_hline(yintercept = 0, color = grey(0.1)) +
geom_pointrange(aes(ymin=ymin, ymax=ymax)
, size = .6
, position = position_dodge(width = 0.4)) +
labs(x = ""
, y = "Confidence interval of WTP estimators (EUR/KG)") +
scale_color_manual(values = c(grey(0.5), grey(0.8))) +
coord_flip()
WTP_plot_pointrange
Kind regards,
Xianwen
--
Xianwen Chen <xnw...@gmail.com>
Graduate student of economics
Norwegian University of Life Sciences
On Tue, May 21, 2013 at 1:05 AM, Dennis Murphy <djmu...@gmail.com> wrote:
> Hi:
> Assignment of objects within data.frame() is rather fraught with
> danger, because they're being assigned in the environment in which the
> data.frame is created. It's safer to use the = syntax inside
> data.frame() because that's the convention for assignment/definition
> inside a function call, but in that case you can't define a new
> variable in terms of another created in the same data.frame() call. It
> needs to be done this way:
>
> rect_MNL_Delta <- data.frame(
> xmin = c(1, 3, 5, 7, 9, 11, 13),
> ymin = c(16.7026, 14.9968, 16.0630, 17.7510, -5.01694,
> -.44146, 1.45884),
> ymax = c(21.1602, 18.7613, 19.1367, 23.6730, 2.26564,
> 3.08630,3.39865)
> )
> # Assign the new variable
> rect_MNL_Delta$xmax <- rect_MNL_Delta$xmin + 1
>
> rect_GMNL <- data.frame(
> xmin = c(1, 3, 5, 7, 9, 11, 13),
> ymin = c(17.20, 16.32, 15.86, 18.12, -8.86, -0.03, 0.95),
> ymax = c(20.12, 18.60, 18.14, 22.29, 3.03, 3.57, 2.81)
> )
> rect_GMNL$xmax <- rect_GMNL$xmin + 1
>
>
> You can fill in the pieces, but I just modified a few of Kaori's
> suggestions to provide a slightly different look. I hate angled text
> on the x-axis, so suggest that you rotate the plot so that these end
> up on the y-axis instead. Since you're using purples and greens, they
> tend to show up better on a white background rather than the default
> gray:
>
> mylabs <- c("Wild cod", "Farmed cod", "Farmed salmon", "Wild monk",
> "Farmed pangasius", "MSC label", "AB label")
>
> ggplot() +
> theme_bw() +
> geom_rect(data = rect_MNL_Delta, aes(xmin = xmin - 0.5,
> xmax = xmax - 0.5,
> ymin = ymin, ymax = ymax),
> fill = "blue") +
> geom_rect(data = rect_GMNL, aes(xmin = xmin - 0.5, xmax = xmax - 0.5,
> ymin = ymin, ymax = ymax),
> fill = "green") +
> scale_x_continuous(breaks = c(1, 3, 5, 7, 9, 11, 13),
> labels = mylabs) +
> coord_flip()
>
> Notice that I subtracted 0.5 from both xmin and xmax so that the bar
> would be centered at the label. I'd suggest making a vector of labels
> outside the ggplot() call and then pull it in during the geom_text()
> call (see the mylabs object and its call within scale_x_continuous()
> ).
>
> Here's another way to approach the problem, both from a data and a
> plotting perspective. In reality, you have a 'fish type' factor; let's
> call it id and redefine the data frames as follows, 'starting from
> scratch':
>
> mylabs <- c("Wild cod", "Farmed cod", "Farmed salmon", "Wild monk",
> "Farmed pangasius", "MSC label", "AB label")
> rect_MNL_Delta <- data.frame(
> id = factor(mylabs, levels = mylabs),
> ymin = c(16.7026, 14.9968, 16.0630, 17.7510, -5.01694,
> -.44146, 1.45884),
> ymax = c(21.1602, 18.7613, 19.1367, 23.6730, 2.26564,
> 3.08630,3.39865)
> )
>
> rect_GMNL <- data.frame(
> id = factor(mylabs, levels = mylabs),
> ymin = c(17.20, 16.32, 15.86, 18.12, -8.86, -0.03, 0.95),
> ymax = c(20.12, 18.60, 18.14, 22.29, 3.03, 3.57, 2.81)
> )
>
> One way to reproduce your plot would be as follows, using
> geom_segment() in place of geom_rect():
>
> ggplot() + theme_bw() +
> geom_segment(data = rect_MNL_Delta,
> aes(x = id, xend = id, y = ymin, yend = ymax),
> size = 10, color = "blue", alpha = 0.5) +
> geom_segment(data = rect_GMNL,
> aes(x = id, xend = id, y = ymin, yend = ymax),
> size = 10, color = "green", alpha = 0.5) +
> labs(x = "", y = "Confidence interval of WTP estimators (EUR/KG)") +
> coord_flip()
>
> Another approach, which I believe would be more appropriate given that
> you want to compare CIs from different models on the same set of
> species, could be done by combining the two data frames and using
> geom_pointrange() on each model to visualize the CIs:
>
> GMNL <- rbind(rect_MNL_Delta, rect_GMNL)
> GMNL$Model <- rep(c("MNL", "GMNL"), each = 7)
> GMNL$mean <- with(GMNL, (ymin + ymax)/2)
>
> ggplot(GMNL, aes(x = id, y = mean, color = Model)) +
> theme_bw() +
> geom_hline(yintercept = 0, color = "gray40") +
> geom_pointrange(aes(ymin = ymin, ymax = ymax),
> size = 1,
> position = position_dodge(width = 0.4)) +
> labs(x = "", y = "Confidence interval of WTP estimators (EUR/KG)") +
> scale_color_manual(values = c("blue", "orange")) +
> coord_flip()
>
> It's not necessarily 'better' than your original plot, but it's an
> alternative approach that allows for direct comparison of the two CIs
> for each id/species/whatever it is ;)
>
> HTH,
> Dennis
>
> On Mon, May 20, 2013 at 1:27 PM, Xianwen Chen <xnw...@gmail.com> wrote:
> > Hi Kaori,
> >
> > ã©ãããããã¨ãï¼ You helped me greatly. I spent two days digging
> > into numerous
> web
> > pages and documents. It was wonderful to get your hints! Just one
> question,
> > why did you use '=' instead of '<-' inside 'data.frame()'?
> >
> > Based on your suggestion, I added Y-axis label and also legends. Because
> > this is a plot for a research article, I have to use black, white, and
> gray.
> > In the following codes, I use two scales of gray. The problem kicks when
> the
> > darker gray totally covers the lighter gray, it is impossible to see
> where
> > the lighter gray is. To clearly illustrate the problem, I put on an
> > identical plot, but with colors at http://i.imgur.com/aehb0dx.jpg
> > Here is the code with colors in gray scales:
> >
> >
> > rect_MNL_Delta <- data.frame(
> > xmin = c(1, 3, 5, 7, 9, 11, 13),
> > xmax = c(2, 4, 6, 8, 10, 12, 14),
> >
> > ymin = c(16.7026, 14.9968, 16.0630, 17.7510, -5.01694, -.44146, 1.45884),
> > ymax = c(21.1602, 18.7613, 19.1367, 23.6730, 2.26564, 3.08630,3.39865),
> > f = 'MNL'
> >
> > )
> > rect_GMNL <- data.frame(
> > xmin = c(1, 3, 5, 7, 9, 11, 13),
> > xmax = c(2, 4, 6, 8, 10, 12, 14),
> >
> > ymin = c(17.20, 16.32, 15.86, 18.12, -8.86, -0.03, 0.95),
> > ymax = c(20.12, 18.60, 18.14, 22.29, 3.03, 3.57, 2.81),
> > f = 'GMNL'
> > )
> >
> > WTP_plot = ggplot() +
> > geom_rect(data = rect_MNL_Delta, aes(xmin = xmin, xmax = xmax, ymin =
> ymin,
> > ymax = ymax, fill = grey(0.3)), alpha = 0.5) +
> > geom_rect(data = rect_GMNL, aes(xmin = xmin, xmax = xmax, ymin = ymin,
> ymax
> > = ymax, fill = grey(0.6)), alpha = 0.5) +
> > scale_fill_manual(name = "Model", breaks = c(grey(0.3), grey(0.6)),
> values
> > =c(grey(0.3), grey(0.6)), labels = c("MNL", "GMNL")) +
> > scale_x_continuous(breaks=c(1.5, 3.5, 5.5, 7.5, 9.5, 11.5, 13.5),
> > labels=c("Wild cod", "Farmed cod","Farmed salmon","Wild monk","Farmed
> > pangasius","MSC label","AB label")) +
> > scale_y_continuous(name = "Confidence Interval of WTP estimators
> (EUR/KG)")
> > +
> > theme(axis.text.x = element_text(angle=45, hjust=.5, vjust=0.5))
> >
> > WTP_plot
> >
> > Is there some way to address this issue? Perhaps same color but with
> > different texture, or pattern? Any thought on this would be highly
> > appreciated. Thank you!
> >
> > Kind regards,
> >
> > Xianwen
> >
> >
> >
> > Ito, Kaori (Groton) wrote:
> >>
> >> How about this? I changed "<-" to "=" in the data.frame.
> >> You have continuous variables on x so I used scale_x_continuous for
> >> changing the ticks and adding labels. You can modify angle, size, etc
> with
> >> theme.
> >>
> >>
> >> rect_MNL_Delta<- data.frame(
> >> xmin = c(1, 3, 5, 7, 9, 11, 13),
> >> xmax = xmin + 1,
> >> ymin = c(16.7026, 14.9968, 16.0630, 17.7510, -5.01694,
> -.44146,
> >> 1.45884),
> >> ymax = c(21.1602, 18.7613, 19.1367, 23.6730, 2.26564,
> >> 3.08630,3.39865)
> >> )
> >> rect_GMNL<- data.frame(
> >> xmin = c(1, 3, 5, 7, 9, 11, 13),
> >> xmax = xmin + 1,
> >> ymin = c(17.20, 16.32, 15.86, 18.12, -8.86, -0.03, 0.95),
> >> ymax = c(20.12, 18.60, 18.14, 22.29, 3.03, 3.57, 2.81)
> >> )
> >>
> >> ggplot() +
> >> geom_rect(data = rect_MNL_Delta, aes(xmin = xmin, xmax = xmax,
> >> ymin = ymin, ymax = ymax), fill = "blue", alpha = 0.1) +
> >> geom_rect(data = rect_GMNL, aes(xmin = xmin, xmax = xmax,
> ymin =
> >> ymin, ymax = ymax), fill = "green", alpha = 0.1) +
> >> scale_x_continuous(breaks=c(1, 3, 5, 7, 9, 11, 13),
> >> labels=c("Farmed cod","text","text","text","text","text","text")) +
> >> theme(axis.text.x = element_text(angle=90, hjust=1,
> vjust=0.5))
> >>
> >>
> >>
> >> -----Original Message-----
> >> From: ggpl...@googlegroups.com [mailto:ggpl...@googlegroups.com] On
> Behalf
> >> Of Xianwen Chen
> >> Sent: Monday, May 20, 2013 11:35 AM
> >> To: ggpl...@googlegroups.com
> >> Subject: Re: [R] bar plot with non-zero starting level
> >>
> >> Hi,
> >>
> >> I want to overlap two rectangular plots. Here is my code:
> >>
> >> require(ggplot2)
> >> rect_MNL_Delta<- data.frame(
> >> xmin<- c(1, 3, 5, 7, 9, 11, 13),
> >> xmax<- xmin + 1,
> >> ymin<- c(16.7026, 14.9968, 16.0630, 17.7510, -5.01694,
> >> -.44146, 1.45884),
> >> ymax<- c(21.1602, 18.7613, 19.1367, 23.6730, 2.26564, 3.08630,
> >> 3.39865)
> >> )
> >> rect_GMNL<- data<- data.frame(
> >> xmin<- c(1, 3, 5, 7, 9, 11, 13),
> >> xmax<- xmin + 1,
> >> ymin<- c(17.20, 16.32, 15.86, 18.12, -8.86, -0.03, 0.95),
> >> ymax<- c(20.12, 18.60, 18.14, 22.29, 3.03, 3.57, 2.81)
> >> )
> >>
> >> ggplot() +
> >> geom_rect(data = rect_MNL_Delta, aes(xmin = xmin, xmax = xmax,
> >> ymin = ymin, ymax = ymax), fill = "blue", alpha = 0.1) +
> >> geom_rect(data = rect_GMNL, aes(xmin = xmin, xmax = xmax, ymin
> >> = ymin, ymax = ymax), fill = "green", alpha = 0.1)
> >>
> >> The problem is that the second geom_rect() fully covered the first
> >> geom_rect(), instead of overlapping with transparency.
> >>
> >> Another issue is that I don't know how to set labels on x-axis, instead
> >> of the current ticks. I would like to have for instance 'Farmed cod'
> >> under the first rectangle.
> >>
> >> Can you provide me some hints? Thank you.
> >>
> >> Kind regards,
> >>
> >> Xianwen
> >>
> >> Jim Lemon wrote:
> >>>
> >>> On 05/19/2013 09:19 AM, Xianwen Chen wrote:
> >>>>
> >>>> Hi,
> >>>>
> >>>> I want to plot grouped bars to compare 95% confidence interval
> estimates
> >>>> from two models. Each bar represents a 95% confidence interval
> estimate
> >>>> of a coefficient from one of the two models. Each group represents
> >>>> confidence interval estimates of the same coefficient from the two
> >>>> models.
> >>>>
> >>>> I think such a bar plot will nicely present whether 95% confidence
> >>>> interval estimates of the same coefficient from the two models
> overlap.
> >>>> All these confidence intervals do not start from the x axis.
> >>>>
> >>>> I searched bar plot examples on Google. I found methods to plot bars
> in
> >>>> groups by barplot(). I could only specify one offset for all
> confidence
> >>>> interval estimates of the same model. I could not individually specify
> >>>> an offset for each confidence interval estimate. Can someone please
> help
> >>>> me on how I may proceed with individual offset for each of the bar in
> >>>> the plot?
> >>>>
> >>> Hi Xianwen,
> >>> It seems that you want to line up either one or both of the estimates
> >>> for each coefficient and then display the confidence intervals as bars
> >>> around the estimates. If so, your ordinate (y-axis) won't be very
> >>> interpretable unless you add one for each pair of coefficients. What I
> >>> would initially suggest is that you display the coefficient values
> >>> along a horizontal line and use "error bars" to show the confidence
> >>> intervals. Here is an example:
> >>>
> >>> coef<-c(3.6,3.2,5.7,6.0,1.2,1.3)
> >>> CIs<-c(1.2,1.4,2.7,2.6,3.1,2.9)
> >>> plot(0,xlim=c(0.5,3.5),ylim=range(c(-CIs,CIs)),type="n",
> >>> xlab="Coefficients",ylab="95% confidence intervals",xaxt="n")
> >>> axis(1,at=c(0.8,1.2,1.8,2.2,2.8,3.2),
> >>> labels=c("A1","A2","B1","B2","C1","C2"))
> >>> abline(h=0)
> >>> library(plotrix)
> >>> dispersion(c(0.8,1.2,1.8,2.2,2.8,3.2),rep(0,6),ulim=CIs,interval=TRUE)
> >>> boxed.labels(c(0.8,1.2,1.8,2.2,2.8,3.2),0,coef)
> >>>
> >>> Jim
> >
> >
> > --
> > --
> > You received this message because you are subscribed to the ggplot2
> mailing
> > list.
> > Please provide a reproducible example:
> > https://github.com/hadley/devtools/wiki/Reproducibility
> >
> > To post: email ggpl...@googlegroups.com
> > To unsubscribe: email ggplot2+unsubscr...@googlegroups.com
> > More options: http://groups.google.com/group/ggplot2
> >
> > --- You received this message because you are subscribed to the Google
> > Groups "ggplot2" group.
> > To unsubscribe from this group and stop receiving emails from it, send an
> > email to ggplot2+unsubscr...@googlegroups.com.
> > For more options, visit https://groups.google.com/groups/opt_out.
> >
> >
>
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
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.
