distribution-analysis

Perform distribution analysis on heavy-tailed distributed data

This repository contains data science tools and methods, and my take and notes on some of them

Investigated whether Fandango's movie ratings were biased by comparing displayed ratings with underlying scores and competing platforms including IMDb, Rotten Tomatoes, and Metacritic. Used Python, Pandas, and Matplotlib to perform exploratory analysis, visualise rating distributions, and evaluate a published data journalism claim.

Python/Seaborn script drawing a facet grid (small multiples) of bathymetric depth distributions across profiles. Reusable scientific-plotting example.

R script drawing a notched box-and-whisker plot of bathymetric depth distributions across 25 trench profiles with ggpubr::ggboxplot (median-confidence notches, magma palette, outlier marking). Mariana Trench; Fig. 6 in Lemenkova (2019), Geodesy and Cartography 45(2):57-84.

Python/Seaborn script drawing kernel-density-estimate (KDE) curve subplots of depth distributions, demonstrated on Mariana Trench data. Reusable scientific-plotting example.

This project aims to generate insights from the sample datasets which are provided.The interest is mainly about gaining insights regarding click-out distribution and click-through rates (CTR).

This is a custom-built library I created, offering practical tools for data analysis and statistical exploration.

End-to-end B2B Sales Performance Dashboard built in Power BI using DAX, Star Schema modeling, and advanced data analysis to uncover revenue drivers and business insights.

Patient Information Analysis for Multi Category Classification Problem

R script drawing kernel-density distribution curves of bathymetric depths, grouped by tectonic plate and by cross-section profile, via data.table melt reshaping and ggplot2 geom_density. Mariana Trench; density analysis in Lemenkova (2019), Geodesy and Cartography 45(2):57-84.

This project aims to generate insights from the sample datasets of sim card online shop. The interest is mainly about gaining insights regarding click-out distribution and click-through rates (CTR).

R scripts building per-profile depth-distribution histograms (ggplot2 geom_histogram with fitted normal and kernel-density curves and mean/median lines) for the 25 Mariana Trench profiles. Fig. 4 in Lemenkova (2019), Geodesy and Cartography 45(2):57-84.

R scripts drawing violin plots (mirrored kernel-density estimates hybridised with box plots) of bathymetric depth distributions across trench profiles, via ggplot2 geom_violin with quantile lines and count scaling and data.table wide-to-long reshaping. Mariana Trench; Lemenkova (2018), Turkish J. Maritime and Marine Sciences 4(2):146-155.

A computational model and a corresponding algorithm to estimate the fibre velocity distribution.

Python script drawing quantile-quantile (Q-Q) plot subplots to check distributional assumptions with statsmodels, on Mariana Trench data. Reusable statistical-graphics example.

R scripts drawing quantile-quantile (Q-Q) and empirical-CDF distributional plots of Mariana Trench depths for 25 profiles, facet-wrapped as small multiples (ggplot2). Fig. 6 in Lemenkova (2019), Geodesy and Cartography 45(2):57-84.

Python/Seaborn script drawing letter-value (boxen) plots of depth distributions, demonstrated on Mariana Trench bathymetry. Reusable scientific-plotting example.

Python/Seaborn script drawing violin plots (box plot, kernel density and median combined) of bathymetric depth distributions, demonstrated on Mariana Trench data. Reusable scientific-plotting example.

Learn how to build a Violin Chart using the Syncfusion .NET MAUI Toolkit to visualize airline arrival delay distributions with medians, IQR, and outlier highlights