1. Data Cleaning

Our approach meticulously filters out anomalies to present a more accurate depiction of population movement patterns. This cleansing is essential to distill patterns, discerning genuine activity from noise, and ensure that our analysis captures authentic population movements.

In examining the aggregated movement data, we identify stationary trends and flows by assessing the duration of presence in specific areas. We record extended gaps in these trends as separate events. This allows us to accurately quantify the length of each activity period and ensures our insights into population movements are precise, capturing the natural flow of activity without extraneous detail.

This refined aggregated data provides a robust foundation for understanding the dynamics of population movement, revealing the rhythms and patterns that characterize how communities interact with their environments.

2. Scaling Factors Computing

Calculate area- and time-specific adjustment ratios based on varied sampling frequencies (observed versus census data).

Raw “true” visits tallies (post-cleansing) x scaling factors = normalized counts. These scaling factors are determined for each region and on a monthly basis. A weighting factor is assigned to each micro-area, calculated in relation to the primary nighttime geolocation detected in the data.

Last updated