https://www.imf.org/external/np/fin/data/rms_mth.aspx?SelectDate=2026-01-01&reportType=REP
These rates, normally quoted as currency units per U.S. dollar, are reported daily to the Fund by the issuing central bank. Rates are normally reported for members whose currencies are used in Fund financial transactions (see Financial Transaction Plan).
In this dataset, I will filter the data to include only the United States and practice working with wide-format data.
https://globalsearch.cuny.edu/CFGlobalSearchTool/CFSearchToolController
From this website, we can easily see that 19 graduate classes are offered. We can copy and paste this untidy dataset into Excel and save it as a CSV file.
These 3 datasets represent real-life data that we have obtained. We will first perform some exploratory data analysis (EDA) and then extract useful results based on the requirements. For the IMF dataset, I will select a few currencies against the USD to observe trends. The Central Bank dataset is a real example of wide-format data. The last dataset, the SPS class schedule, may not be considered wide-format, but it is an example of untidy data.
library(untidydata)
data("language_diversity")
head(language_diversity) Continent Country Measurement Value
1 Africa Algeria Langs 18
2 Africa Angola Langs 42
3 Oceania Australia Langs 234
4 Asia Bangladesh Langs 37
5 Africa Benin Langs 52
6 Americas Bolivia Langs 38library(tidyr)
library(dplyr)
Attaching package: 'dplyr'The following objects are masked from 'package:stats':
filter, lagThe following objects are masked from 'package:base':
intersect, setdiff, setequal, uniondf_ld_wide <- pivot_wider(language_diversity,
names_from = Measurement,
values_from = Value)
df_ld_wide# A tibble: 74 × 8
Continent Country Langs Area Population Stations MGS Std
<chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 Africa Algeria 18 2381741 25660 102 6.6 2.29
2 Africa Angola 42 1246700 10303 50 6.22 1.87
3 Oceania Australia 234 7713364 17336 134 6 4.17
4 Asia Bangladesh 37 143998 118745 20 7.4 0.73
5 Africa Benin 52 112622 4889 7 7.14 0.99
6 Americas Bolivia 38 1098581 7612 48 6.92 2.5
7 Africa Botswana 27 581730 1348 10 4.6 1.69
8 Americas Brazil 209 8511965 153322 245 9.71 5.87
9 Africa Burkina Faso 75 274000 9242 6 5.17 1.07
10 Africa CAR 94 622984 3127 13 8.08 1.21
# ℹ 64 more rowsdf_ld_long <- df_ld_wide %>%
mutate(ContinentCode = case_when(
Continent == "Africa" ~ 1,
Continent == "Asia" ~ 2,
Continent == "Oceania" ~ 3,
Continent == "Americas" ~ 4
)) %>%
select(-Continent)
df_ld_long <- df_ld_long %>%
pivot_longer(cols = -Country, names_to = "Metric", values_to = "Value")
df_ld_long# A tibble: 518 × 3
Country Metric Value
<chr> <chr> <dbl>
1 Algeria Langs 18
2 Algeria Area 2381741
3 Algeria Population 25660
4 Algeria Stations 102
5 Algeria MGS 6.6
6 Algeria Std 2.29
7 Algeria ContinentCode 1
8 Angola Langs 42
9 Angola Area 1246700
10 Angola Population 10303
# ℹ 508 more rows# df_ld_long %>%
# group_by(ContinentCode) %>% # group by continent
# summarise(count = n()) # df_ld_long <- df_ld_long %>%
# mutate(ContinentCode = case_when(
# Continent == "Africa" ~ 1,
# Continent == "Asia" ~ 2,
# Continent == "Oceania" ~ 3,
# Continent == "Americas" ~ 4
#
# )) %>%
# select(-Continent)
#
# df_ld_longNow Africa=1, Asia=2, Oceania=3, Americas=4.
df_ld_reshape <- df_ld_long %>%
pivot_wider(names_from =Country, values_from = Value)
df_ld_reshape# A tibble: 7 × 75
Metric Algeria Angola Australia Bangladesh Benin Bolivia Botswana Brazil
<chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 Langs 1.8 e1 4.2 e1 2.34e2 37 5.2 e+1 3.8 e1 2.7 e1 2.09e2
2 Area 2.38e6 1.25e6 7.71e6 143998 1.13e+5 1.10e6 5.82e5 8.51e6
3 Population 2.57e4 1.03e4 1.73e4 118745 4.89e+3 7.61e3 1.35e3 1.53e5
4 Stations 1.02e2 5 e1 1.34e2 20 7 e+0 4.8 e1 1 e1 2.45e2
5 MGS 6.6 e0 6.22e0 6 e0 7.4 7.14e+0 6.92e0 4.6 e0 9.71e0
6 Std 2.29e0 1.87e0 4.17e0 0.73 9.9 e-1 2.5 e0 1.69e0 5.87e0
7 Continent… 1 e0 1 e0 3 e0 2 1 e+0 4 e0 1 e0 4 e0
# ℹ 66 more variables: `Burkina Faso` <dbl>, CAR <dbl>, Cambodia <dbl>,
# Cameroon <dbl>, Chad <dbl>, Colombia <dbl>, Congo <dbl>,
# `Costa Rica` <dbl>, `Cote d'Ivoire` <dbl>, Cuba <dbl>, Ecuador <dbl>,
# Egypt <dbl>, Ethiopia <dbl>, `French Guiana` <dbl>, Gabon <dbl>,
# Ghana <dbl>, Guatemala <dbl>, Guinea <dbl>, Guyana <dbl>, Honduras <dbl>,
# India <dbl>, Indonesia <dbl>, Kenya <dbl>, Laos <dbl>, Liberia <dbl>,
# Libya <dbl>, Madagascar <dbl>, Malawi <dbl>, Malaysia <dbl>, Mali <dbl>, …With the dataset in wide format, comparing two countries becomes straightforward and easy for the end user to read. However, if we want to plot any country in R, we still need to reshape the data back to a longer format. For example, I selected Algeria and Costa Rica for comparison.
df_2_compare <- df_ld_reshape %>%
select(Metric,Algeria, `Costa Rica`)
df_2_compare <-df_2_compare %>%
pivot_longer(cols = c(Algeria, `Costa Rica`),
names_to = "Country",
values_to = "Value")
df_2_compare# A tibble: 14 × 3
Metric Country Value
<chr> <chr> <dbl>
1 Langs Algeria 18
2 Langs Costa Rica 10
3 Area Algeria 2381741
4 Area Costa Rica 51100
5 Population Algeria 25660
6 Population Costa Rica 3064
7 Stations Algeria 102
8 Stations Costa Rica 38
9 MGS Algeria 6.6
10 MGS Costa Rica 8.92
11 Std Algeria 2.29
12 Std Costa Rica 1.78
13 ContinentCode Algeria 1
14 ContinentCode Costa Rica 4 library(tidyverse)── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
✔ forcats 1.0.1 ✔ readr 2.1.6
✔ ggplot2 4.0.1 ✔ stringr 1.6.0
✔ lubridate 1.9.4 ✔ tibble 3.3.1
✔ purrr 1.2.1
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag() masks stats::lag()
ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errorsggplot(df_2_compare, aes(x = Metric, y = Value, fill = Country)) +
geom_col(position = "dodge") + # side-by-side bars
labs(title = "Comparison of Metrics: Algeria vs Costa Rica",
x = "Metric",
y = "Value") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
# df_long %>%
# group_by(Country, Metric) %>%
# summarise(count = n()) %>%
# filter(count > 1)df_long <- df_ld_reshape %>%
pivot_longer(
cols = -Metric, # all columns except Metric
names_to = "Country", # new column storing country names
values_to = "Value" # temporary column for metric values
) %>%
pivot_wider(
names_from = Metric, # turn metric names into columns
values_from = Value # fill cells with the values
)
df_long# A tibble: 74 × 8
Country Langs Area Population Stations MGS Std ContinentCode
<chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 Algeria 18 2381741 25660 102 6.6 2.29 1
2 Angola 42 1246700 10303 50 6.22 1.87 1
3 Australia 234 7713364 17336 134 6 4.17 3
4 Bangladesh 37 143998 118745 20 7.4 0.73 2
5 Benin 52 112622 4889 7 7.14 0.99 1
6 Bolivia 38 1098581 7612 48 6.92 2.5 4
7 Botswana 27 581730 1348 10 4.6 1.69 1
8 Brazil 209 8511965 153322 245 9.71 5.87 4
9 Burkina Faso 75 274000 9242 6 5.17 1.07 1
10 CAR 94 622984 3127 13 8.08 1.21 1
# ℹ 64 more rowsHere, we separate each metric indicator into multiple columns to provide more detailed insights for every indicator.
ggplot(df_fileter <- df_long %>% filter(Country %in% c("Algeria", "Costa Rica")), aes(x = Country, y = Langs, fill = Country)) +
geom_col(position = "dodge") + # side-by-side bars
labs(title = "Comparison of Langs: Algeria vs Costa Rica",
x = "Langs",
y = "Value") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
url <- "https://raw.githubusercontent.com/dyc-sps/-dyc-sps-SPS_Data607_Week6/refs/heads/main/IMF_currencies_exchange_rate.csv"
df_imf_rate <- read.csv(url)
head(df_imf_rate) Currency X2.Jan.26 X5.Jan.26 X6.Jan.26 X7.Jan.26 X8.Jan.26 X9.Jan.26
1 Chinese yuan <NA> 6.9806 6.9803 6.9913 6.983 6.982
2 Euro(1) 1.1721 1.1664 1.1707 1.1684 1.1675 1.1642
3 Japanese yen <NA> 157.24 156.67 156.61 156.77 157.06
4 U.K. pound(1) 1.34375 1.346 1.35215 1.3502 1.34455 1.34125
5 U.S. dollar 1 1 1 1 1 1
6 Algerian dinar <NA> 129.788 129.5562 129.7706 129.8228 129.9849
X12.Jan.26 X13.Jan.26 X14.Jan.26 X15.Jan.26 X16.Jan.26 X20.Jan.26 X21.Jan.26
1 6.9746 6.9764 6.9732 6.9714 6.9683 6.9606 6.9638
2 1.1692 1.1654 1.1651 1.1624 1.1617 1.1728 1.1739
3 <NA> 158.24 159.2 158.5 158.2 158.15 158.21
4 1.346 1.34715 1.34495 1.34185 1.34085 1.3463 1.34085
5 1 1 1 1 1 1 1
6 <NA> 129.9154 129.9966 130.0776 130.153 129.7996 129.6896
X22.Jan.26 X23.Jan.26 X26.Jan.26 X27.Jan.26 X28.Jan.26 X29.Jan.26 X30.Jan.26
1 6.9633 6.9647 6.9577 6.9571 6.9454 6.9469 6.9488
2 1.1706 1.1742 1.1836 1.1929 1.1974 1.1968 1.1919
3 158.2 158.55 154.77 154.29 152.58 153.09 153.6
4 1.34215 1.35165 1.3667 1.37115 1.37705 1.37925 1.37795
5 1 1 1 1 1 1 1
6 129.7822 129.5853 129.3217 129.229 129.1428 129.1078 129.2009To plot the Chinese Yuan against the Japanese Yen from this dataset, I first need to reshape it into long format and filter the relevant countries before creating the plot.
df_imf_rate <- df_imf_rate %>%
pivot_longer(
cols = -Currency, # all columns except Currency
names_to = "IMF_Date", # new column storing IMF_Date names
values_to = "C_Rate" ) # temporary column for C_Rate values
# ) %>%
#
# pivot_wider(
# names_from = Metric, # turn metric names into columns
# values_from = Value # fill cells with the values
# )
df_imf_rate$IMF_Date <- sub("^X", "",df_imf_rate$IMF_Date)
df_imf_rate$IMF_Date <- as.Date(df_imf_rate$IMF_Date, format = "%d.%b.%y")
df_imf_rate# A tibble: 720 × 3
Currency IMF_Date C_Rate
<chr> <date> <chr>
1 Chinese yuan 2026-01-02 <NA>
2 Chinese yuan 2026-01-05 6.9806
3 Chinese yuan 2026-01-06 6.9803
4 Chinese yuan 2026-01-07 6.9913
5 Chinese yuan 2026-01-08 6.983
6 Chinese yuan 2026-01-09 6.982
7 Chinese yuan 2026-01-12 6.9746
8 Chinese yuan 2026-01-13 6.9764
9 Chinese yuan 2026-01-14 6.9732
10 Chinese yuan 2026-01-15 6.9714
# ℹ 710 more rowsggplot(df_fileter <- df_imf_rate %>% filter(Currency %in% c("Chinese yuan", "Japanese yen"))%>% drop_na(), aes(x = IMF_Date, y = C_Rate, group=Currency, color=Currency,fill = Currency)) +
#geom_col(position = "dodge") + # side-by-side bars
geom_line() +
labs(title = "Jan 2026 - Chinese yuan vs Japanese yen",
x = "Date",
y = "C_Rate") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
For this chart, it’s necessary to re-sort the dates and apply a log transformation to the differences, since the exchange rates vary widely. The Chinese Yuan generally ranges from 6 to 7, whereas the Japanese Yen is around 150.
options(max.print = 40) # to control total number of values printed.
url2<- "https://raw.githubusercontent.com/dyc-sps/-dyc-sps-SPS_Data607_Week6/refs/heads/main/IMF_CentralBankData_US.csv"
df_us_cbs <- read.csv(url2)
head(df_us_cbs) DATASET SERIES_CODE OBS_MEASURE COUNTRY INDICATOR TYPE_OF_TRANSFORMATION
FREQUENCY SCALE DECIMALS_DISPLAYED SECTOR MFS_COLTN COUNTERPART_SECTOR
MFS_INSTRL MFS_RA MFS_AGGREGATES ACCOUNTING_ENTRY FI_MATURITY CURRENCY
VALUATION MFS_SRF MFS_COMPONENT FR_ADJ EXRATE TRANSFORMATION UNIT OVERLAP
MFS_EAWR IFS_FLAG DOI FULL_DESCRIPTION AUTHOR PUBLISHER DEPARTMENT
CONTACT_POINT TOPIC TOPIC_DATASET KEYWORDS KEYWORDS_DATASET LANGUAGE
PUBLICATION_DATE
[ reached 'max' / getOption("max.print") -- omitted 506 columns ]
[ reached 'max' / getOption("max.print") -- omitted 6 rows ]This is a wide dataset with over 500 columns.
options(max.print = 40)
colnames(df_us_cbs) [1] "DATASET" "SERIES_CODE" "OBS_MEASURE"
[4] "COUNTRY" "INDICATOR" "TYPE_OF_TRANSFORMATION"
[7] "FREQUENCY" "SCALE" "DECIMALS_DISPLAYED"
[10] "SECTOR" "MFS_COLTN" "COUNTERPART_SECTOR"
[13] "MFS_INSTRL" "MFS_RA" "MFS_AGGREGATES"
[16] "ACCOUNTING_ENTRY" "FI_MATURITY" "CURRENCY"
[19] "VALUATION" "MFS_SRF" "MFS_COMPONENT"
[22] "FR_ADJ" "EXRATE" "TRANSFORMATION"
[25] "UNIT" "OVERLAP" "MFS_EAWR"
[28] "IFS_FLAG" "DOI" "FULL_DESCRIPTION"
[31] "AUTHOR" "PUBLISHER" "DEPARTMENT"
[34] "CONTACT_POINT" "TOPIC" "TOPIC_DATASET"
[37] "KEYWORDS" "KEYWORDS_DATASET" "LANGUAGE"
[40] "PUBLICATION_DATE"
[ reached 'max' / getOption("max.print") -- omitted 506 entries ]df_us_cbs <- df_us_cbs %>%
# mutate(across(where(is.numeric), as.character))
mutate(across(everything(), as.character))
df_us_cbs DATASET SERIES_CODE OBS_MEASURE COUNTRY INDICATOR TYPE_OF_TRANSFORMATION
FREQUENCY SCALE DECIMALS_DISPLAYED SECTOR MFS_COLTN COUNTERPART_SECTOR
MFS_INSTRL MFS_RA MFS_AGGREGATES ACCOUNTING_ENTRY FI_MATURITY CURRENCY
VALUATION MFS_SRF MFS_COMPONENT FR_ADJ EXRATE TRANSFORMATION UNIT OVERLAP
MFS_EAWR IFS_FLAG DOI FULL_DESCRIPTION AUTHOR PUBLISHER DEPARTMENT
CONTACT_POINT TOPIC TOPIC_DATASET KEYWORDS KEYWORDS_DATASET LANGUAGE
PUBLICATION_DATE
[ reached 'max' / getOption("max.print") -- omitted 506 columns ]
[ reached 'max' / getOption("max.print") -- omitted 6 rows ]df_us_cbs <- df_us_cbs %>%
pivot_longer(
cols = -c(COUNTRY,SERIES_CODE), # all columns except COUNTRY
names_to = "CBS_Date", # new column storing IMF_Date names
values_to = "Amount" ) # temporary column for C_Rate values
df_us_cbs# A tibble: 3,264 × 4
SERIES_CODE COUNTRY CBS_Date Amount
<chr> <chr> <chr> <chr>
1 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States DATASET IMF.S…
2 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States OBS_MEASURE OBS_V…
3 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States INDICATOR Liabi…
4 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States TYPE_OF_TRANSFORMATI… US do…
5 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States FREQUENCY Month…
6 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States SCALE Milli…
7 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States DECIMALS_DISPLAYED <NA>
8 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States SECTOR <NA>
9 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States MFS_COLTN <NA>
10 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States COUNTERPART_SECTOR <NA>
# ℹ 3,254 more rowslibrary(stringr)
df_filtered <- df_us_cbs %>%
#filter(CBS_Date %in% c("USA.S121_L_LT_S1311MIXED_CBS.USD.M", "USA.S121_L_MB_CBS.USD.M")) %>%
filter(str_detect(CBS_Date, "^X\\d{4}\\.M\\d{2}$")) %>%
drop_na()
df_filtered <- df_filtered %>%
mutate(Amount = as.numeric(Amount))
print(df_filtered)# A tibble: 570 × 4
SERIES_CODE COUNTRY CBS_Date Amount
<chr> <chr> <chr> <dbl>
1 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States X2001.M12 6645
2 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States X2002.M01 13688
3 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States X2002.M02 5752
4 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States X2002.M03 5692
5 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States X2002.M04 5387
6 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States X2002.M05 5883
7 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States X2002.M06 8116
8 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States X2002.M07 6242
9 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States X2002.M08 4874
10 USA.S121_L_LT_S1311MIXED_CBS.USD.M United States X2002.M09 7879
# ℹ 560 more rowsggplot(df_filtered, aes(x = CBS_Date, y = Amount, color =SERIES_CODE,group=SERIES_CODE, fill = SERIES_CODE)) +
#geom_col(position = "dodge") + # side-by-side bars
geom_line() +
labs(title = "Liabilities to Central Government (CBS) vs Liabilities, Monetary base (CBS) Monthly",
x = "Date",
y = "Millions") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
ggsave("df_us_cbs_line_plot.png",
plot = last_plot(), # your last ggplot
width = 10, # width in inches
height = 6, # height in inches
dpi = 300) df_filtered <- df_us_cbs %>%
filter(str_detect(CBS_Date, "^X\\d{4}\\.Q\\d{1}$")) %>%
drop_na()
#print(df_filtered)
df_filtered <- df_filtered %>%
mutate(Amount = as.numeric(Amount))
ggplot(df_filtered, aes(x = CBS_Date, y = Amount, color =SERIES_CODE,group=SERIES_CODE, fill = SERIES_CODE)) +
#geom_col(position = "dodge") + # side-by-side bars
geom_line() +
labs(title = "Liabilities to Central Government (CBS) vs Liabilities, Monetary base (CBS) Quarterly",
x = "Date",
y = "Millions") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
df_filtered <- df_us_cbs %>%
filter(str_detect(CBS_Date, "^X\\d{4}$")) %>%
drop_na()
#print(df_filtered)
df_filtered <- df_filtered %>%
mutate(Amount = as.numeric(Amount))
ggplot(df_filtered, aes(x = CBS_Date, y = Amount, color =SERIES_CODE,group=SERIES_CODE, fill = SERIES_CODE)) +
#geom_col(position = "dodge") + # side-by-side bars
geom_line() +
labs(title = "Liabilities to Central Government (CBS) vs Liabilities, Monetary base (CBS) Annual",
x = "Date",
y = "Millions") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
Wide and long formats each have their strengths. Wide format is easier for humans to read and quickly compare values across multiple variables, making it ideal for tables and reports. Long format, on the other hand, is better suited for data processing, analysis, and plotting in tools, because it allows for easier filtering, grouping, and reshaping. In practice, it’s common to switch between the two formats depending on whether the focus is on readability or data manipulation.
• OpenAI. (2025). ChatGPT (Version 5.2) [Large language model]. https://chat.openai.com. Accessed Mar 08, 2026.