Skip to content

Project Overview.md

Latest commit

 

History

History
71 lines (46 loc) · 2.69 KB

Project Overview.md

File metadata and controls

71 lines (46 loc) · 2.69 KB

Project Overview: Analyzing Google Trends Data

Step-by-Step Process

  1. Data Collection

Objective: Collect Google Trends data for specific climate change-related keywords across different cities.

  • Tools Used: pytrends library for accessing Google Trends data.
  • Process:
    • Defined a list of keywords related to climate change for each city.
    • Used pytrends to fetch data for each keyword and city over a specified timeframe.
    • Normalized the data using population size and internet penetration rates to ensure comparability across cities.

Reason: Normalization accounts for differences in population and internet access, providing a more accurate representation of search interest.

  1. Temporal Analysis

Objective: Analyze temporal trends in the collected data.

  • Tools Used: pandas, matplotlib, statsmodels.
  • Process:
    • Performed STL decomposition to separate long-term trends, seasonal components, and residual noise.
    • Created line graphs for each city, annotated with key global events related to climate change.

Reason: STL decomposition helps identify underlying patterns in time series data, while annotations provide context for observed trends.

  1. Geographical Analysis

Objective: Compare search interest across cities and assess spatial patterns.

  • Tools Used: pandas, seaborn, geopandas, pysal.
  • Process:
    • Generated Kernel Density Estimation (KDE) plots to visualize search interest density.
    • Created heatmaps to show regional search intensity.
    • Calculated Moran’s I to assess spatial autocorrelation.

Reason: These analyses help understand how search interest varies geographically and whether it is spatially clustered.

  1. Visualization

Objective: Create visualizations for presentation.

  • Tools Used: matplotlib, seaborn.
  • Process:
    • Generated line graphs, heatmaps, and KDE plots.
    • Ensured visuals were labeled and annotated with major events.

Reason: Visualizations provide a clear and engaging way to communicate findings.

  1. Statistical Analysis

Objective: Perform statistical tests to validate findings.

  • Tools Used: pymannkendall, libpysal, esda.
  • Process:
    • Conducted the Mann-Kendall Test to detect monotonic trends in temporal data.
    • Used Moran’s I to assess spatial autocorrelation.

Reason: Statistical tests provide quantitative evidence for observed trends and patterns.

  1. Documentation and Reporting

Objective: Document the process and results.

  • Tools Used: Markdown for report writing.
  • Process:
    • Created a detailed markdown report summarizing the methodology, results, and implications.
    • Included recommendations based on findings.

Reason: Documentation ensures transparency and provides a reference for stakeholders to understand and act on the findings.