InR-evolutionbyFatih Tüzen·Nov 3Introducing tuiksdmx: A Metadata-Driven R Client for Turkish Official StatisticsDiscover datasets, explore SDMX metadata, and retrieve official statistics from Türkiye using R.
InR-evolutionbyFatih Tüzen·Aug 13Why ACF Is More Than Just a PlotWhat autocorrelation reveals — and sometimes hides — about how a time series remembers its past
InR-evolutionbyFatih Tüzen·May 8Differencing: A Transformation or a Trap?How differencing reshapes persistence, signal, and interpretation in time series analysis
Fatih Tüzen·Apr 24Publishing a Quarto Blog: What I Learned Moving from Netlify to GitHub PagesQuarto makes it surprisingly easy to build a blog.
InR-evolutionbyFatih Tüzen·Apr 17Why Most Time Series Models Fail Before They StartA practical look at stationarity and transformations with real CPI data in RA response icon2A response icon2
Fatih Tüzen·Feb 11Statistics Is Not About Numbers, It’s About TrustMost people think statistics is about numbers. Percentages, averages, growth rates, rankings.This belief is understandable — and deeply…
InR-evolutionbyFatih Tüzen·Jan 22Data Leakage in R: Why Correct Evaluation Matters Even When Metrics Do Not ChangeThis article examines data leakage as a methodological problem rather than a performance issue. Using a real-world dataset and repeated…
InR-evolutionbyFatih Tüzen·Jan 2Data Normalization in R: When, Why, and How to Scale Your Data CorrectlyNormalization is not a technical afterthought but a modeling decision. When, why, and how you scale your data can change model behavior —…A response icon1A response icon1
Fatih Tüzen·Dec 26, 2025Understanding Data Import and Export in R: Working with CSV and Excel FilesI work with data where getting it wrong at the beginning can quietly ruin everything that follows.
InR-evolutionbyFatih Tüzen·Dec 19, 2025Outliers in Data Analysis: Detecting Extreme Values Before Modeling in R with İstanbul Airbnb DataOutliers are often treated as errors to be removed.In real-world data analysis, many extreme values are meaningful signals rather than…A response icon1A response icon1