Reputation: 6657
I want to draw a line graph where there is a clear outlier. The following dataset is a subset of the one I'm using:
Date PCROI
717 2014-06-28 38
718 2014-06-29 41
719 2014-06-30 36
720 2014-07-01 734
721 2014-07-02 60
722 2014-07-03 78
723 2014-07-04 90
724 2014-07-05 68
725 2014-07-06 55
726 2014-07-07 57
If I plot this using ggplot2
the graph is not that bad..
ggplot(data = example, aes(x = Date, y = PCROI)) + geom_line()
Unfortunately the time series I'm working with is much longer than the median of PCROI is 51. How can I adapt the chart without excluding the value from the dataset?
Upvotes: 1
Views: 2124
Reputation: 1363
You can use the subset
command if you have some criteria for outliers (e.g., if PRCOI is greater than 100) and if you're fine with interpolating the deleted point:
example <- read.table(header = TRUE, text = " Date PCROI
717 2014-06-28 38
718 2014-06-29 41
719 2014-06-30 36
720 2014-07-01 734
721 2014-07-02 60
722 2014-07-03 78
723 2014-07-04 90
724 2014-07-05 68
725 2014-07-06 55
726 2014-07-07 57")
example$Date <- as.Date(example$Date)
ggplot(subset(example, PCROI <= 100), aes(Date, PCROI)) + geom_line()
Upvotes: 1