Data Science Stack Exchange is a question and answer site for Data science professionals, Machine Learning specialists, and those interested in learning more about the field. It only takes a minute to sign up.
Sign up to join this community
I have a time series data that has some missing dates. For example, in the dataset below '2017-08-06' is missing.
To enter this data in the Prophet model, do I have to create the data frame with all the missing dates? I know the Prophet model does not require the missing data for any specific time to be filled as it is a curve-fitting-based model. But I am unsure if all dates should be included for daily time series prediction. Thanks so much for answering my question.
If you are going to work exclusively in Prophet it is robust in the absence of a reasonable number of missing serial data points.
If you intend to do traditional time series work as well, yes, you will need to have a space for each regular interval and ideally values. Traditional time series models use subsets of values at lags to estimate future time values. When there are missing values, there is nothing to predict from. So finding a method appropriate to you specific data, uses and the level of and reason for missingness is imperative!
