Work with Time Series Containing Missing Data 

Thanks to the symbolic nature of Mathematica the values of time series may be any expressions. In particular, the expression Missing[…] is reserved to annotate missing observations, and Mathematica provides a special handling for it.

The time series of temperatures readings in Champaign, Illinois on May 14, 2014.

In[1]:=

X wd = WeatherData[{"Champaign", "IL"}, "Temperature", {2014, 5, 14}]

Out[1]=

In[2]:=

X Cases[wd["Path"], {t_, mv_Missing} :> {DateObject[t], mv}]

Out[2]=

Out[940]=

The missing values are treated as symbolic expressions.

In[3]:=

X Mean[wd]

Out[3]=

Use MissingDataMethod option to specify how to fill in the missing value.

In[4]:=

X ts1 = TimeSeries[wd, MissingDataMethod -> "Interpolation"];

In[5]:=

X ts2 = TimeSeries[wd, MissingDataMethod -> {"Constant", Quantity[10.7, "DegreeCelsius"]}];

Compare means.

In[6]:=

X {Mean[ts1], Mean[ts2]}

Out[6]=