问题
Facebook's Prophet in R (there's also a Python version) is used to generate time series forecasts.
A model m
is created by:
m <- prophet(df)
future <- make_future_dataframe(m, periods = 365)
forecast <- predict(m, future)
plot(m, forecast)
Which returns a very nicely formatted graph, like:
I would like to change the line type, to get not dots but a usual thin line.
I had tried this
lines(m$history$y,lty=1)
but got an error
In doTryCatch(return(expr), name, parentenv, handler)
Are there are any suggestions how to convert those dots into a line?
回答1:
The plot
method for prophet
objects uses ggplot2
, so base R graphics functions like lines()
won't work. You can use ggplot2::geom_line()
to add lines, but at the moment I don't see an easy way to replace the points by lines ...
Example from ?prophet
:
history <- data.frame(ds = seq(as.Date('2015-01-01'), as.Date('2016-01-01'), by = 'd'),
y = sin(1:366/200) + rnorm(366)/10)
m <- prophet(history)
future <- make_future_dataframe(m, periods = 365)
forecast <- predict(m, future)
pp <- plot(m,forecast)
Add lines:
library(ggplot2)
pp + geom_line()
This question provides a (hacky) way forward:
pp2 <- pp + geom_line()
qq2 <- ggplot_build(pp2)
qq2$data[[2]]$colour <- NA
plot(ggplot_gtable(qq2))
But obviously something went wrong with the hack. The better bet would be to look at the plot method(prophet:::plot.prophet
) and modify it to behave as you want ... Here is the bare-bones version:
df <- prophet:::df_for_plotting(m, forecast)
gg <-ggplot(df, aes(x = ds, y = y)) + labs(x = "ds", y = "y")
gg <- gg + geom_ribbon(ggplot2::aes(ymin = yhat_lower,
ymax = yhat_upper), alpha = 0.2, fill = "#0072B2",
na.rm = TRUE)
## replace first geom_point() with geom_line() in next line ...
gg <- gg + geom_line(na.rm = TRUE) + geom_line(aes(y = yhat),
color = "#0072B2", na.rm = TRUE) + theme(aspect.ratio = 3/5)
I may have stripped out some components that exist in your data/forecast, though ...
回答2:
it is possible to make such manipulations with dyplot.prophet(m, forecast)
(html version of plot) :) before that, we should rewrite function like here:
dyplot.prophet <- function(x, fcst, uncertainty=TRUE,
...)
{
forecast.label='Predicted'
actual.label='Actual'
# create data.frame for plotting
df <- prophet:::df_for_plotting(x, fcst)
# build variables to include, or not, the uncertainty data
if(uncertainty && exists("yhat_lower", where = df))
{
colsToKeep <- c('y', 'yhat', 'yhat_lower', 'yhat_upper')
forecastCols <- c('yhat_lower', 'yhat', 'yhat_upper')
} else
{
colsToKeep <- c('y', 'yhat')
forecastCols <- c('yhat')
}
# convert to xts for easier date handling by dygraph
dfTS <- xts::xts(df %>% dplyr::select_(.dots=colsToKeep), order.by = df$ds)
# base plot
dyBase <- dygraphs::dygraph(dfTS)
presAnnotation <- function(dygraph, x, text) {
dygraph %>%
dygraphs::dyAnnotation(x, text, text, attachAtBottom = TRUE)
}
dyBase <- dyBase %>%
# plot actual values
dygraphs::dySeries(
'y', label=actual.label, color='black',stepPlot = TRUE, strokeWidth=1
) %>%
# plot forecast and ribbon
dygraphs::dySeries(forecastCols, label=forecast.label, color='blue') %>%
# allow zooming
dygraphs::dyRangeSelector() %>%
# make unzoom button
dygraphs::dyUnzoom()
if (!is.null(x$holidays)) {
for (i in 1:nrow(x$holidays)) {
# make a gray line
dyBase <- dyBase %>% dygraphs::dyEvent(
x$holidays$ds[i],color = "rgb(200,200,200)", strokePattern = "solid")
dyBase <- dyBase %>% dygraphs::dyAnnotation(
x$holidays$ds[i], x$holidays$holiday[i], x$holidays$holiday[i],
attachAtBottom = TRUE)
}
}
return(dyBase)
}
the strokeWidth=0
was before and we have changed it to strokeWidth=1
and added stepPlot = TRUE
the whole basis code is situated here: https://rdrr.io/cran/prophet/src/R/plot.R
来源:https://stackoverflow.com/questions/53947623/how-to-change-type-of-line-in-prophet-plot