In MatthewHeun/IEATools: Tools for Working with International Energy Agency Data

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
library(dplyr)
library(IEATools)
library(tidyr)

Introduction

After loading IEA data and ensuring that the data are balanced (see the vignette about loading data) and specifying many of the flows (see the vignette about specifying flows), the analyst can prepare for physical supply-use table (PSUT) analysis of an energy conversion chain. This vignette shows how to use the functions in the IEATools package for that purpose.

Adding PSUT matrix names (add_psut_matnames())

The first step to converting the IEA extended energy balance data into PSUT matrices is identifying the matrix in which each datum belongs. The add_psut_matnames() function performs this task. It adds a column named "matnames" (by default but changeable by the user) that contains one of R, U, V, or Y to identify the Resources, Use, Make, or Final demand matrix.

load_tidy_iea_df() %>% 
  add_psut_matnames() %>%
  glimpse()

Note that when a matnames column already exists in .tidy_iea_df, add_psut_matnames() returns the input data frame without modification. This behavior allows other matrix naming conventions.

Add matrix metadata (add_row_col_meta())

The next step is to add information to identify the rows and columns for each datum. The add_row_col_meta() function performs this task by adding rowname, colname, rowtype, and coltype columns which identify the names and types of rows and columns in the PSUT matrices. add_row_col_meta() should be called after calling add_psut_matnames(), because it relies on the the matnames column to determine names and types.

load_tidy_iea_df() %>% 
  add_psut_matnames() %>%
  add_row_col_meta() %>% 
  glimpse()

Note that if all of rowname, colname, rowtype, or coltype columns are already present in .tidy_iea_data, add_row_col_meta() returns the input data frame without modification. This behavior allows other matrix row, column, rowtype, and coltype naming conventions.

If some but not all of rowname, colname, rowtype, and coltype columns are present in in .tidy_iea_data, add_row_col_meta()returns an error.

Collapse to matrices (collapse_to_tidy_psut())

After identifying row and column names and types, the next step is to collapse the data into PSUT matrices. The collapse_to_tidy_psut() function performs this task using the matsindf::collapse_to_matrices() function internally. The matvals column is added to the data frame and contain matrix names for each row of the data frame. collapse_to_tidy_psut() ensures that all entries in each matrix are non-negative.

Tidy_PSUT <- 
load_tidy_iea_df() %>% 
  specify_all() %>% 
  add_psut_matnames() %>% 
  add_row_col_meta() %>% 
  collapse_to_tidy_psut()
Tidy_PSUT %>% 
  glimpse()
class(Tidy_PSUT$matvals[[1]])
Tidy_PSUT$matvals[[1]]

Generate S_units matrices (extract_S_units_from_tidy())

A matrix of Products and Units can be obtained from a tidy IEA data frame with the extract_S_units_from_tidy() function.

S_units_df <- load_tidy_iea_df() %>% 
  specify_all() %>% 
  extract_S_units_from_tidy()
S_units_df %>% 
  glimpse()
S_units_df$S_units[[1]]

Putting it all together (prep_psut())

prep_psut() is a convenience function. It converts a tidy IEA data frame that has gone through specify_all() into a PSUT data frame by collapsing the IEA data into PSUT matrices (R, U, V, and Y).

This function bundles several others: 1. add_psut_matnames() 2. add_row_col_meta() 3. collapse_to_tidy_psut()

Furthermore, it extracts S_units matrices using extract_S_units_from_tidy() and adds those matrices to the data frame.

Simple <- load_tidy_iea_df() %>% 
  specify_all() %>% 
  prep_psut() %>% 
  pivot_longer(cols = c(R, U_EIOU, U_feed, V, Y, S_units), names_to = "matnames", values_to = "matvals")
S_units <- load_tidy_iea_df() %>% 
  specify_all() %>% 
  extract_S_units_from_tidy()
Complicated <- load_tidy_iea_df() %>% 
  specify_all() %>% 
  add_psut_matnames() %>% 
  add_row_col_meta() %>% 
  collapse_to_tidy_psut() %>% 
  spread(key = matnames, value = matvals) %>% 
  full_join(S_units, by = c("Method", "Energy.type", "Last.stage", 
                            "Country", "Year")) %>% 
  gather(key = matnames, value = matvals, R, U_EIOU, U_feed, 
         V, Y, S_units)
# Simple and Complicated are same.
full_join(Simple %>% rename(matval_simple = matvals), 
          Complicated %>% rename(matval_complicated = matvals), 
          by = c("Method", "Energy.type", "Last.stage", "Country", "Year", "matnames")) %>% 
  dplyr::mutate(
    same = matsbyname::equal_byname(matval_simple, matval_complicated)
  ) %>% 
  magrittr::extract2("same") %>% 
  as.logical() %>% 
  all()
Simple %>% 
  glimpse()
# U_EIOU for Ghana in 1971
Simple$matvals[[5]]

Conclusion

Taken together, the functions in this package make it possible to prepare IEA extended energy balance data for use with PSUT analysis techniques. A sequence of three function calls is all that is required.

wide_PSUT_mats <- load_tidy_iea_df() %>% 
  specify_all() %>% 
  prep_psut()
wide_PSUT_mats %>% 
  glimpse()

MatthewHeun/IEATools documentation built on Feb. 6, 2024, 3:29 p.m.