Use non-parametric methods to calculate time series trends
smoothTrend(
mydata,
pollutant = "nox",
deseason = FALSE,
type = "default",
statistic = "mean",
avg.time = "month",
percentile = NA,
data.thresh = 0,
simulate = FALSE,
n = 200,
autocor = FALSE,
cols = "brewer1",
shade = "grey95",
xlab = "year",
y.relation = "same",
ref.x = NULL,
ref.y = NULL,
key.columns = length(percentile),
name.pol = pollutant,
ci = TRUE,
alpha = 0.2,
date.breaks = 7,
auto.text = TRUE,
k = NULL,
plot = TRUE,
...
)an openair object
A data frame containing the field date and at least one
other parameter for which a trend test is required; typically (but not
necessarily) a pollutant.
The parameter for which a trend test is required. Mandatory.
Should the data be de-deasonalized first? If TRUE the
function stl is used (seasonal trend decomposition using loess).
Note that if TRUE missing data are first imputed using a
Kalman filter and Kalman smooth.
type determines how the data are split i.e. conditioned,
and then plotted. The default is will produce a single plot using the
entire data. Type can be one of the built-in types as detailed in
cutData e.g. “season”, “year”, “weekday” and
so on. For example, type = "season" will produce four plots --- one
for each season.
It is also possible to choose type as another variable in the data
frame. If that variable is numeric, then the data will be split into four
quantiles (if possible) and labelled accordingly. If type is an existing
character or factor variable, then those categories/levels will be used
directly. This offers great flexibility for understanding the variation of
different variables and how they depend on one another.
Type can be up length two e.g. type = c("season", "weekday") will
produce a 2x2 plot split by season and day of the week. Note, when two
types are provided the first forms the columns and the second the rows.
Statistic used for calculating monthly values. Default is
“mean”, but can also be “percentile”. See timeAverage
for more details.
Can be “month” (the default), “season” or “year”. Determines the time over which data should be averaged. Note that for “year”, six or more years are required. For “season” the data are plit up into spring: March, April, May etc. Note that December is considered as belonging to winter of the following year.
Percentile value(s) to use if statistic =
"percentile" is chosen. Can be a vector of numbers e.g. percentile
= c(5, 50, 95) will plot the 5th, 50th and 95th percentile values
together on the same plot.
The data capture threshold to use (%) when aggregating
the data using avg.time. A value of zero means that all available
data will be used in a particular period regardless if of the number of
values available. Conversely, a value of 100 will mean that all data will
need to be present for the average to be calculated, else it is recorded
as NA. Not used if avg.time = "default".
Should simulations be carried out to determine the
Mann-Kendall tau and p-value. The default is FALSE. If TRUE,
bootstrap simulations are undertaken, which also account for
autocorrelation.
Number of bootstrap simulations if simulate = TRUE.
Should autocorrelation be considered in the trend uncertainty
estimates? The default is FALSE. Generally, accounting for
autocorrelation increases the uncertainty of the trend estimate sometimes
by a large amount.
Colours to use. Can be a vector of colours e.g. cols =
c("black", "green") or pre-defined openair colours --- see
openColours for more details.
The colour used for marking alternate years. Use “white” or “transparent” to remove shading.
x-axis label, by default “year”.
This determines how the y-axis scale is plotted. "same"
ensures all panels use the same scale and "free" will use panel-specific
scales. The latter is a useful setting when plotting data with very
different values. ref.x See ref.y for details. In this case the
correct date format should be used for a vertical line e.g. ref.x =
list(v = as.POSIXct("2000-06-15"), lty = 5).
See ref.y.
A list with details of the horizontal lines to be added
representing reference line(s). For example, ref.y = list(h = 50,
lty = 5) will add a dashed horizontal line at 50. Several lines can be
plotted e.g. ref.y = list(h = c(50, 100), lty = c(1, 5), col =
c("green", "blue")). See panel.abline in the lattice
package for more details on adding/controlling lines.
Number of columns used if a key is drawn when using the
option statistic = "percentile".
Names to be given to the pollutant(s). This is useful if you want to give a fuller description of the variables, maybe also including subscripts etc.
Should confidence intervals be plotted? The default is
FALSE.
The alpha transparency of shaded confidence intervals - if plotted. A value of 0 is fully transparent and 1 is fully opaque.
Number of major x-axis intervals to use. The function
will try and choose a sensible number of dates/times as well as formatting
the date/time appropriately to the range being considered. This does not
always work as desired automatically. The user can therefore increase or
decrease the number of intervals by adjusting the value of
date.breaks up or down.
Either TRUE (default) or FALSE. If
TRUE titles and axis labels will automatically try and format
pollutant names and units properly e.g. by subscripting the ‘2’ in
NO2.
This is the smoothing parameter used by the gam function in
package mgcv. By default it is not used and the amount of smoothing
is optimised automatically. However, sometimes it is useful to set the
smoothing amount manually using k.
Should a plot be produced? FALSE can be useful when
analysing data to extract plot components and plotting them in other
ways.
Other graphical parameters are passed onto cutData and
lattice:xyplot. For example, smoothTrend passes the option
hemisphere = "southern" on to cutData to provide southern
(rather than default northern) hemisphere handling of type =
"season". Similarly, common graphical arguments, such as xlim and
ylim for plotting ranges and pch and cex for plot
symbol type and size, are passed on xyplot, although some local
modifications may be applied by openair. For example, axis and title
labelling options (such as xlab, ylab and main) are
passed to xyplot via quickText to handle routine formatting.
One special case here is that many graphical parameters can be vectors
when used with statistic = "percentile" and a vector of
percentile values, see examples below.
David Carslaw
The smoothTrend function provides a flexible way of estimating the
trend in the concentration of a pollutant or other variable. Monthly mean
values are calculated from an hourly (or higher resolution) or daily time
series. There is the option to deseasonalise the data if there is evidence
of a seasonal cycle.
smoothTrend uses a Generalized Additive Model (GAM) from the
gam package to find the most appropriate level of smoothing.
The function is particularly suited to situations where trends are not
monotonic (see discussion with TheilSen() for more details on
this). The smoothTrend function is particularly useful as an
exploratory technique e.g. to check how linear or non-linear trends are.
95% confidence intervals are shown by shading. Bootstrap estimates of the
confidence intervals are also available through the simulate option.
Residual resampling is used.
Trends can be considered in a very wide range of ways, controlled by setting
type - see examples below.
Other time series and trend functions:
TheilSen(),
calendarPlot(),
runRegression(),
timePlot(),
timeProp(),
timeVariation(),
trendLevel()
# trend plot for nox
smoothTrend(mydata, pollutant = "nox")
# trend plot by each of 8 wind sectors
if (FALSE) smoothTrend(mydata, pollutant = "o3", type = "wd", ylab = "o3 (ppb)")
# several pollutants, no plotting symbol
if (FALSE) smoothTrend(mydata, pollutant = c("no2", "o3", "pm10", "pm25"), pch = NA)
# percentiles
if (FALSE) smoothTrend(mydata, pollutant = "o3", statistic = "percentile",
percentile = 95)
# several percentiles with control over lines used
if (FALSE) smoothTrend(mydata, pollutant = "o3", statistic = "percentile",
percentile = c(5, 50, 95), lwd = c(1, 2, 1), lty = c(5, 1, 5))
Run the code above in your browser using DataLab