Thoughts & Code
It can happen that some categorical features have lot of different values with small frequencies. This can often be a problem during modeling phase. Typical solution is to merge all categories with small (e.g. 5% , 1% etc… ) frequency of appearance in data set into one category e.g. Other Below is the pandas example […]
Often we need to change sampling of a time series data set that we are working with. Imagine that you are working with time series that was sampled in 15 minutes chunks and need to do analysis/ forecast with different sampling e.g. daily , 3 hours, quarterly, yearly etc… There is method resample in pandas […]
In this post I will try to explain how to extract top frequencies from the time series in python. It is a useful feature that helps with time series analysis, time series decomposition , forecasting etc… I will try to focus on those topics in next few posts. Loading Sorry, something went wrong. Reload? Sorry, […]