Table of Contents: :: Library Catalog

Saved in:

Bibliographic Details
Main Author:	AlessandroSpike
Format:	Recurso digital
Language:
Published:	Zenodo 2025
Online Access:	https://doi.org/10.5281/zenodo.15188174
Tags:	Add Tag No Tags, Be the first to tag this record!

Table of Contents:

This folders contains files to reproduce results of the paper "Carbon trade biases and the emerging mesoscale structure of the European Emissions Trading System network" by Andrea Flori and Alessandro Spelta. %% GRAVITY ESTIMATION File name: gravity_model.R This file reproduces the results reported in Table 1 and in Supplementary Information S1. The code is based on R. The analysis requires a set of R packages to be installed. They are reported in the first 38 lines. Line 46: we import the main .csv file, where relevant information is located. Lines 49-51: we create the Same Country (C) variable Lines 54-69: we create the Same Sector (S) variable for each year (as a function of aggregate verified emissions of the installations in the sample in each NUTS-3 province) Lines 73-215: for each year separately, the code computes the gravity model estimation using the package ppml. Specifically, the y variable is the amount of traded allowances in each year. The geodesic distance is labelled "distance". The pedix "acq" refers to the destination node, while the pedix "transf" to the origin node. We fit Eq. 1 of the paper. Lines 219-509: we construct summary statistics for the estimates of the gravity models. These estimates are conveniently saved in the object "stats_db". Lines 513-535: we create the database for the plots. Lines 538-735: these lines build and export all the plots reported in the paper for the main analysis. Lines 740-790: we calculate the deviance and the fitted values for each model. %% PLOTS AND COMMUNITY DETECTION ANALYSIS %BiasCountrySectorReal This script calculates three types of bias (country, sector, country-sector) in both weighted and binary network connections over time. Bias is measured as percentage of within-group connections relative to total connections. Input required: AdjMatrixTime.mat containing adjacency matrices (Adj) and timestamps (Tempi), plus partition vectors for countries and sectors. Output: BiasReal.mat containing bias calculations over time. %BiasCountrySectorFitted This script analyzes network bias in gravity model outputs across three dimensions: country, sector, and country-sector combinations. For each dimension, both weighted and binary (unweighted) bias metrics are calculated. Input required: OutputGravity.mat: Contains gravity model results (iso_d, iso_o, fitted, year) AdjMatrixTime.mat: Contains partition vectors for countries and sectors Output: BiasFitted.mat: Contains bias calculations (biasFitted) for each time period (Tempi) The bias is calculated as percentage of within-group connections relative to total connections, for both weighted and unweighted network representations. %DistanceAnalysisReal This script analyzes trade/interaction patterns based on geographical distance and country borders using real data. It processes trade amounts between different locations and analyzes them across distance quartiles and borders. Input required: OutputGravity.mat: Contains trade data (iso_d, iso_o, amount, year) AdjMatrixTime.mat: Contains time periods, location codes, and distance matrix Border_Distance_Info: Contains border information and location names Output: ConfiniReal.mat: Contains analyzed trade patterns by distance quartiles and border regions %DistanceAnalysisFitted This script analyzes trade/interaction patterns based on geographical distance and country borders using fitted values from a gravity model. It processes fitted interaction values between different locations and analyzes them across distance quartiles and borders. Input required: OutputGravity.mat: Contains fitted gravity model values and ISO codes AdjMatrixTime.mat: Contains time periods, location codes, and distance matrix Border_Distance_Info: Contains border information and location names Output: ConfiniFitted.mat: Contains analyzed interaction patterns by distance quartiles and border regions %PlotBias This script creates visualizations comparing real and fitted network characteristics, including bias analysis, border effects, and distance-based weight distributions. Input required: BiasReal.mat: Contains bias measures from real network BiasFitted.mat: Contains bias measures from fitted network ConfiniFitted.mat: Contains analyzed fitted patterns by distance and borders ConfiniReal.mat: Contains analyzed real patterns by distance and borders Outputs: Figure 1: Bias comparison and border region analysis Figure 2: Distance-based weight distribution analysis %GeoNetPlot This script visualizes the European Union Emissions Trading System (EU ETS) network using geographical data and adjacency matrices. Input: NUTS_RG_60M_2021_3035.shp: Shape file containing EU regional boundaries AdjMatrixAggregate.mat: Adjacency matrix representing connections between regions Output: Geographical visualization of the EU ETS network showing: Nodes: Regions sized by their connection strength Edges: Connections between regions Colors: Based on outgoing connection strength (logarithmic scale) %PlotDeviance This MATLAB script analyzes and visualizes deviance data for different gravity models of carbon emission trading. It compares various model configurations and calculates confidence intervals for the difference in deviances. Input: CSV files containing deviance data for different model configurations Years range from 2005 to 2020 Output: Two figures: Deviance plots for different model configurations Log of deviance difference with confidence intervals Console output of difference in deviances and confidence intervals %CommunityOT This script performs a network analysis over time using Optimal Transport (OT) methods and classical modularity approaches. It computes modularity and partition metrics for a set of temporal networks and compares them with community structures based on countries and sectors. Inputs: OutputGravity.mat: Contains data on countries, fitted values, and amounts. AdjMatrixTime.mat: Contains adjacency matrices and partition information. CommunityAgg.mat: Provides aggregated community information. Outputs: Plots and figures representing modularity and partition metrics over time. Saved figures in 'PlotFig/' directory. %CommunityOT_AggregateNet This script analyzes transportation and cultural distance data across different regions over time. It uses optimal transport methods to assess community interactions based on given data. The script outputs plots showing average cultural distances within and between communities. Inputs: 'OutputGravity.mat': Contains: iso_d: Destination identifiers. iso_o: Origin identifiers. amount: Amount of transport or interaction. fitted: Fitted values from a gravity model. year: Corresponding years for the data. 'AdjMatrixTime.mat': Contains: Tempi: Time periods for the analysis. C: List of community identifiers. Distance: Matrix of distances (not used in this script). 'Border_Distance_Info.mat': Contains: Nomi: Border information (not used in this script). DummyBorder: Dummy variables for border information (not used). 'culturaldistance.mat': Contains: ctr: Country codes. Various columns of cultural distance data. Outputs: Figures showing the average cultural distances within and between communities based on different factors. Optional: Saves community aggregation data to 'CommunityAgg.mat'.

Similar Items