I have some data which contains data going back to 2009.
However, for each individual in the dataset they may have 10 or so data points related to them. At infrequent and varied intervals between each data point.
Also, with their being 600 or so individuals in the dataset, each individual doesn't have many data points associated to them.
So is there a forecasting library/method in Python or R which can help me predict the value for a certain reading?
On the one hand, you can forecast each series separately. There is little you can do for time series with only ten observations. You can try the historical mean, or the historical median, or use the last historical observation as a forecast. No special tools or libraries are necessary. See Best method for short time-series.
On the other hand, you may be able to do some kind of cross-learning, so you can learn from the other 599 individuals when you forecast individual #600. There are a number of ways of doing this, e.g., fitting a mixed linear model or possibly setting up an appropriate neural network (although 10 observations each on 600 individuals is again not a lot of data for a neural network). You would need to provide more information about your specific use case.
