Ukraine War: Russian Losses Timeline

Author

Cormac Nolan

Published

2022-05-21

Import

Data sourced from the Kyiv Independent Twitter feed who sourced it from the Ukrainian Defense Ministry.

These are imperfect estimates, given during chaotic wartime conditions, from an institution that is would be incentivised to inflate its figures. This isn’t to say that the Defense Ministry is consciously inflating them, but it bears remembering when looking at this data.

# A tibble: 78 x 14
   report_date soldiers planes helicopters tanks artillery_pieces
   <date>         <dbl>  <dbl>       <dbl> <dbl>            <dbl>
 1 2022-02-23         0      0           0     0                0
 2 2022-02-26      3500     14           8   102               15
 3 2022-02-27      4300     27          26   146               49
 4 2022-02-28      5300     29          29   191               74
 5 2022-03-01      5710     29          29   198               77
 6 2022-03-02      5840     30          31   211               85
 7 2022-03-03      9000     30          31   217               90
 8 2022-03-04      9166     33          37   251              105
 9 2022-03-05     10000     39          40   269              105
10 2022-03-06     11000     44          48   285              109
# ... with 68 more rows, and 8 more variables:
#   armoured_personel_carriers <dbl>, mlrs <dbl>, boats <dbl>, vehicles <dbl>,
#   fuel_tankers <dbl>, uav <dbl>, aa_systems <dbl>, cruise_missiles <dbl>

Rates

First we need to apply interpolation in case there were any missed days, linear is fine.

 [1] "2022-02-23" "2022-02-24" "2022-02-25" "2022-02-26" "2022-02-27"
 [6] "2022-02-28" "2022-03-01" "2022-03-02" "2022-03-03" "2022-03-04"
[11] "2022-03-05" "2022-03-06" "2022-03-07" "2022-03-08" "2022-03-09"
[16] "2022-03-10" "2022-03-11" "2022-03-12" "2022-03-13" "2022-03-14"
[21] "2022-03-15" "2022-03-16" "2022-03-17" "2022-03-18" "2022-03-19"
[26] "2022-03-20" "2022-03-21" "2022-03-22" "2022-03-23" "2022-03-24"
[31] "2022-03-25" "2022-03-26" "2022-03-27" "2022-03-28" "2022-03-29"
[36] "2022-03-30" "2022-03-31" "2022-04-01" "2022-04-02" "2022-04-03"
[41] "2022-04-04" "2022-04-05" "2022-04-06" "2022-04-07" "2022-04-08"
[46] "2022-04-09" "2022-04-10" "2022-04-11" "2022-04-12" "2022-04-13"
[51] "2022-04-14" "2022-04-15" "2022-04-16" "2022-04-17" "2022-04-18"
[56] "2022-04-19" "2022-04-20" "2022-04-21" "2022-04-22" "2022-04-23"
[61] "2022-04-24" "2022-04-25" "2022-04-26" "2022-04-27" "2022-04-28"
[66] "2022-04-29" "2022-04-30" "2022-05-01" "2022-05-02" "2022-05-03"
[71] "2022-05-04" "2022-05-05" "2022-05-06" "2022-05-07" "2022-05-08"
[76] "2022-05-09" "2022-05-10" "2022-05-11" "2022-05-12" "2022-05-13"
[81] "2022-05-14" "2022-05-15" "2022-05-16" "2022-05-17" "2022-05-18"
[86] "2022-05-19" "2022-05-20" "2022-05-21"

Now let’s calculate rates…

# A tibble: 5 x 3
  report_date soldiers tanks
  <date>         <dbl> <dbl>
1 2022-02-24      1167    34
2 2022-02-25      1166    34
3 2022-02-26      1167    34
4 2022-02-27       800    44
5 2022-02-28      1000    45

Important Events

# A tibble: 1 x 5
  report_date event_name     event_category event_location source
  <date>      <chr>          <chr>          <chr>          <chr> 
1 2022-04-03  Moskva Sinking naval          Black Sea      <NA>

Rate Plots Timeline

APC’s - April Onwards

Tanks - April Onward

Aircraft

Artillery

--- title: "Ukraine War: Russian Losses Timeline" author: "Cormac Nolan" date: "`r Sys.Date()`" output: html_document format: html: code-fold: true code-summary: "Show the code" code-tools: true --- ```{r setup, include=FALSE, message=FALSE, error=FALSE} knitr::opts_chunk$set(echo = FALSE) library(dplyr) library(ggplot2) library(forecast) library(here) ``` ## Import Data sourced from the [Kyiv Independent Twitter feed](https://twitter.com/KyivIndependent/media) who sourced it from the Ukrainian Defense Ministry. These are imperfect estimates, given during chaotic wartime conditions, from an institution that is would be incentivised to inflate its figures. This isn't to say that the Defense Ministry is consciously inflating them, but it bears remembering when looking at this data. ```{r import} source(here("data/ukraine_defense_ministry_russian_loss_estimates.R")) campaign_length = as.integer(max(russian_losses$report_date) - min(russian_losses$report_date)) campaign_length_weeks = ceiling(campaign_length / 7) russian_losses ``` ## Rates First we need to apply interpolation in case there were any missed days, linear is fine. ```{r, fig.height=7} russian_losses_all_dates <- dplyr::tibble(report_date = min(russian_losses$report_date) + 0:campaign_length) %>% left_join(russian_losses, by="report_date") russian_losses_all_dates <- russian_losses_all_dates %>% mutate(across(where(~is.numeric(.x)), .fns=forecast::na.interp)) %>% mutate(across(where(~is.numeric(.x)), .fns=as.numeric)) %>% mutate(across(where(~is.numeric(.x)), .fns=round)) russian_losses_all_dates <- russian_losses_all_dates %>% mutate(campaign_week_number = c(0, rep(1:campaign_length_weeks, each=7)[1:(campaign_length)])) russian_losses_all_dates$report_date ``` Now let's calculate rates... ```{r} daily_diff <- function(val){ val - dplyr::lag(val) } calculate_loss_rate <- function(df){ df %>% mutate(across(where(is.numeric), .fns=daily_diff), .keep = "unused") %>% tidyr::drop_na() } russian_losses_rates <- russian_losses_all_dates %>% select(-campaign_week_number) %>% calculate_loss_rate() select(russian_losses_rates, report_date, soldiers, tanks) %>% head(5) ``` ### Important Events ```{r} source(here("./data/ukraine_war_important_events.R")) important_events %>% head(1) ``` ## Rate Plots Timeline #### APC's - April Onwards ```{r} ukraine_blue <- "#0057b7" ukraine_blue_light <- "#ACC8E7" ukraine_blue_very_light <- "#EDF4FA" ukraine_blue_dark <- "#00346d" ukraine_yellow <- "#ffd700" ukraine_yellow_darkened <- "#d6b500" text_colour <- "#555555" ukraine_plot_theme <- function(){ theme(plot.title = element_text(size = 20, colour = text_colour), strip.text.x = element_text(size = 16, colour = text_colour), axis.title = element_text(size = 15, face="bold", colour = text_colour), axis.text.x = element_text( angle = 30, hjust = 1, colour = text_colour, size = 13 ), axis.text.y = element_text(size = 13), panel.background = element_rect( fill = ukraine_blue_very_light, colour = ukraine_blue_light ) ) } format_names <- function(in_string){ out_string <- in_string %>% stringr::str_replace_all("_", " ") %>% stringr::str_to_title() out_string <- ifelse(out_string == "Uav", yes = "UAV", no = out_string) out_string } prepare_rate_plot_data <- function(rates_df, time_col, ...) { rates_df %>% select({{time_col}}, ...) %>% rename_with(.fn=format_names, .cols=-contains("date")) %>% tidyr::pivot_longer( -{{time_col}}, names_to = "category", values_to = "loss_count" ) } rate_plot <- function( rates_df, events_df, category_for_plot, plot_title, plot_x, plot_y ) { current_plot_rates <- rates_df %>% filter(category == category_for_plot) current_plot_rates %>% ggplot(aes(report_date, loss_count)) + geom_bar(stat = "identity", alpha = 0.5, fill = ukraine_yellow_darkened) + geom_smooth( se=FALSE, span=0.7, method = "loess", formula = y ~ x, colour = ukraine_blue, size = 1.5 ) + geom_vline(mapping = aes(xintercept = report_date), data = events_df, colour = ukraine_blue_dark, linetype="dashed", alpha = 0.8, size = 1.1) + geom_text( mapping = aes( x = report_date + 1, label = event_name, y = max(current_plot_rates$loss_count) / 2 ), data = mutate(events_df, category = category_for_plot), colour = ukraine_blue_dark, angle = 90, size = 5 ) + theme_minimal() + ukraine_plot_theme() + labs(title = plot_title, x = plot_x,y = plot_y) } ``` ```{r, fig.width=8, fig.height=6, warning=FALSE, message=FALSE} russian_losses_rates %>% prepare_rate_plot_data( report_date, armoured_personel_carriers ) %>% filter(report_date >= "2022-04-01") %>% rate_plot( events_df = filter( important_events, event_category == "land", report_date >= "2022-04-01" ), category_for_plot = "Armoured Personel Carriers", plot_title = "Russian Daily Loss Rates: APC's", plot_x = "Report Date", plot_y="Loss Count" ) ``` ### Tanks - April Onward ```{r, fig.width=8, fig.height=6, warning=FALSE, message=FALSE} russian_losses_rates %>% prepare_rate_plot_data( report_date, tanks ) %>% filter(report_date >= "2022-04-01") %>% rate_plot( events_df = filter( important_events, event_category == "land", report_date >= "2022-04-01" ), category_for_plot = "Tanks", plot_title = "Russian Daily Loss Rates: Tanks", plot_x = "Report Date", plot_y="Loss Count" ) ``` #### Aircraft ```{r, fig.width=8, fig.height=6, warning=FALSE, message=FALSE} russian_losses_rates %>% prepare_rate_plot_data( report_date, uav ) %>% rate_plot( events_df = filter(important_events, event_category == "land"), category_for_plot = "UAV", plot_title = "Russian Daily Loss Rates: UAV's", plot_x = "Report Date", plot_y="Loss Count" ) ``` ```{r, fig.width=8, fig.height=6, warning=FALSE, message=FALSE} russian_losses_rates %>% prepare_rate_plot_data( report_date, helicopters ) %>% rate_plot( events_df = filter(important_events, event_category == "land"), category_for_plot = "Helicopters", plot_title = "Russian Daily Loss Rates: Helicopters", plot_x = "Report Date", plot_y="Loss Count" ) ``` #### Artillery ```{r, fig.width=8, fig.height=6, warning=FALSE, message=FALSE} russian_losses_rates %>% prepare_rate_plot_data( report_date, artillery_pieces ) %>% rate_plot( events_df = filter(important_events, event_category == "land"), category_for_plot = "Artillery Pieces", plot_title = "Russian Daily Loss Rates: Artillery", plot_x = "Report Date", plot_y="Loss Count" ) ```