Statistical package for non-parametric trend analysis in time series data
sandeep kumar patakamuri (~sandeep_kumar) |
Trend analysis in time series data is classified as parametric and non-parametric methods. Mann-Kendall test and Spearman's Rho test are the two most widely used methods of non-parametric trend detection. Most commonly used non-parametric method like the Mann-Kendall test assumes the data to be serially independent. In the presence of serially correlated data, the tests indicate erroneous results.
To address the serial correlation issue, various modifications to the Mann-Kendall test and Spearman's rho test were suggested in literature like
- Trend-free pre-whitening
- Bias-corrected pre-Whitening
- Variance correction approach
- Block bootstrapping, etc.
Each of these methods has its own advantages and disadvantages. To arrive at a better consensus, it is suggested to use various methods for testing the trends. There is a huge debate in the scientific community to find the optimal method to estimate correct trends in time-series data. Unfortunately, there is no package or library in python to experiment with various approaches suggested in the literature.
An immense need for a library package in python is expressed by researchers from various fields of science. In the current work, we are presenting a python package to perform trend analysis in time series data. The package serves as a platform for analyzing the trends in various fields like meteorology, hydrology, environmental sciences, social sciences, medicine, etc.
- Basic programming skills in python
Sandeep Kumar Patakamuri is a doctoral student from Anna University, Chennai, India. He has been working on Land Cover Land Use Change monitoring and modeling using Earth Observation Satellite Systems for the past seven years. His research interests include climate change modeling, applications of space technology for disaster risk reduction and management, small satellites, digital image processing, and participatory research methods.
He is a very active member of NASA LCLUC South/Southeast Asia Regional Initiative, Global Land Programme and Global Observation of Forest Cover and Land Dynamics (GOFC-GOLD). He is the recipient of a prestigious fellowship from START global initiative and represented India in GOFC-GOLD Advance Data initiative and workshop.
He is an avid traveler, a couch-surfer and likes to socialize during his travels.