Lenguajes de programación: evolución y programación en Economía

Alejandro Lucas, Luis Díaz, Adrián Guijarro

19/12/23

0. Intro

En este trabajo hemos realizado un seguimiento de los aspectos determinantes en la adopción de disitintos lenguajes de programación(popularidad, salarios, lenguajes más utilizados por jóvenes,etc.), así como un estudio sobre la aplicación de la programación para la Economía.

Esto último es fundamental para poner en práctica los conocimientos adquiridos a lo largo del grado y contrastar las teorías y modelos estudiados con la evidencia empírica.

1. Análisis del índice TIOBE

• Lenguajes más populares
• Evolución top 10 lenguajes

Código

library(tiobeindexr)
library(tidyverse)
library(gt)
library(RColorBrewer)
library(ggthemes)
library(highcharter)
library(gtExtras)

df <- top_20()
df <-  df %>% select(1,4,5)
df <- df %>%
  mutate(porcentaje_numeric = ifelse(str_detect(Ratings, "%"),
                                     as.numeric(str_remove(Ratings, "%")),
                                     as.numeric(Ratings))) %>%
  rename(cuota = porcentaje_numeric) %>%
  select(1,2,4)
 

colores <- c("#3498db", "#2ecc71", "#e74c3c", "#f39c12", "#9b59b6", "#e67e22", "#1abc9c", "#95a5a6")

df_logos <- data.frame(lenguaje = c("Python", "C", "C++", "Java", "C#", "JavaScript", "PHP", "Visual Basic", "SQL", "Assembly language", "Scratch", "Fortran", "Go", "MATLAB", "Kotlin", "Delphi/Object Pascal", "Swift","Ruby","R","Rust"),
                       logo = c("https://tse4.mm.bing.net/th?id=OIP.EDJ9xoErBbZqK2tExVoJfAAAAA&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.bkbn2-K7c9rMBV5dvYXDrQHaIh&pid=Api&rs=1&c=1&qlt=95&w=103&h=118","https://tse1.mm.bing.net/th?id=OIP.ggb-3B-8LRfbgS0lPLYNxwHaIU&pid=Api&rs=1&c=1&qlt=95&w=105&h=118","https://tse1.mm.bing.net/th?id=OIP.dJ7A1hPb-BDRBrhOmhrVAgHaHa&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.2x1yN8k3MVCrkJFR_C5k-wHaIp&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.YPYdVAg5ieRspLI6HIpmDQHaIB&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.pkqphAig1t-PCsy4dkVrfAHaD5&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.Y224uoZnCs8yY0Sos_h4cQHaDd&pid=Api&rs=1&c=1&qlt=95&w=258&h=120","https://tse4.mm.bing.net/th?id=OIP.S_9pgiKacbi3FxPKxDmusgHaFy&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.v5m2HW3FI3RbopnK8Cs6HgAAAA&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.bHt6sdtkwUqP91oPKlruGgHaEM&pid=Api&P=0&h=180","https://tse2.explicit.bing.net/th?id=OIP.tLfIejOZiP_H0Zse0NIn2QAAAA&pid=Api&P=0&h=180","https://tse3.mm.bing.net/th?id=OIP.7nKR4oRAhgtSbD97mDXWWgHaE8&pid=Api&P=0&h=180","https://tse2.mm.bing.net/th?id=OIP.cI8KflTKlnmraGFa22XpKQHaEK&pid=Api&P=0&h=180","https://tse3.mm.bing.net/th?id=OIP.gW_YaXsZmuSntd_sy5zkIwHaHa&pid=Api&P=0&h=180","https://tse2.mm.bing.net/th?id=OIP.hEP7hOlvIzhKU1fGpizd7QAAAA&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.XozeB5BRQSzNCO6k7om62QHaHa&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.2BwE10YiTSXRrnW2LnVh8wHaHa&pid=Api&P=0&h=180","https://tse3.mm.bing.net/th?id=OIP.0rg-o4mEzLO5EHJ1zS52YwHaGe&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.YmPfYoQK3WvGEjFXusKV9QHaHa&pid=Api&P=0&h=180"))

df_final <- left_join(df, df_logos, by = c(`Programming Language` = "lenguaje"))
df_final <- df_final %>% mutate(Logo = logo, .after = `Programming Language`)
df_final <- df_final %>% select(-5)

minimal_table_prueba <- gt(df_final) %>% 
  gtExtras::gt_plt_dot(column = cuota, category_column = `Programming Language`,  max_value = NULL, palette = colores ) %>% 
  gtExtras::gt_theme_nytimes() %>% 
  tab_header(title = "Lenguajes de programación más populares",
             subtitle = "Según el índice TIOBE.") %>% 
  cols_width(`Programming Language` ~ px(368), 3 ~ px(30)) %>%
  gtExtras::gt_img_rows(columns = Logo, img_source = 'web')

minimal_table_prueba

Lenguajes de programación más populares
Según el índice TIOBE.
Dec 2023	Programming Language	Logo	cuota
1	Python		13.86
2	C		11.44
3	C++		10.01
4	Java		7.99
5	C#		7.30
6	JavaScript		2.90
7	PHP		2.01
8	Visual Basic		1.82
9	SQL		1.61
10	Assembly language		1.11
11	Scratch		1.08
12	Fortran		1.07
13	Go		1.03
14	MATLAB		0.93
15	Kotlin		0.92
16	Delphi/Object Pascal		0.92
17	Swift		0.82
18	Rust		0.80
19	Ruby		0.77
20	R		0.72

Código

df_2 <- long_term_history()
df_2 <- df_2 %>% pivot_longer(cols = 2:9, names_to = "año", values_to = "posicion")
df_2 <- df_2 %>% mutate(posicion =  as.numeric(posicion),
                        año = as.numeric(año))

df_python <- df_2 %>% filter(`Programming Language` == "Python")
df_c <- df_2 %>% filter(`Programming Language` == "C")
df_cplus <- df_2 %>% filter(`Programming Language` == "C++")
df_cplus <- df_2 %>% filter(`Programming Language` == "C++")
df_java <- df_2 %>% filter(`Programming Language` == "Java")
df_javasc <- df_2 %>% filter(`Programming Language` == "JavaScript")
df_csharp <- df_2 %>% filter(`Programming Language` == "C#")
df_csharp <- df_2 %>% filter(`Programming Language` == "C#")
df_php <- df_2 %>% filter(`Programming Language` == "PHP")
df_vb <- df_2 %>% filter(`Programming Language` == "Visual Basic")
df_sql <- df_2 %>% filter(`Programming Language` == "SQL")
df_asm <- df_2 %>% filter(`Programming Language` == "Assembly language")


df_evolucion <- highchart() %>%
  hc_chart(type = "line", backgroundColor = "white") %>%
  hc_add_series(df_python, "line", hcaes(año, posicion), name = "Python") %>%
  hc_add_series(df_c, "line", hcaes(año, posicion), name = "C") %>%
  hc_add_series(df_cplus, "line", hcaes(año, posicion), name = "C++") %>%
  hc_add_series(df_java, "line", hcaes(año, posicion), name = "Java") %>%
  hc_add_series(df_javasc, "line", hcaes(año, posicion), name = "JavaScript") %>%
  hc_add_series(df_csharp, "line", hcaes(año, posicion), name = "C#") %>%
  hc_add_series(df_php, "line", hcaes(año, posicion), name = "PHP") %>%
  hc_add_series(df_vb, "line", hcaes(año, posicion), name = "Visual Basic") %>%
  hc_add_series(df_sql, "line", hcaes(año, posicion), name = "SQL") %>%
  hc_add_series(df_asm, "line", hcaes(año, posicion), name = "Assembly Language") %>%
  hc_title(text = "Evolución del Top 10 lenguajes de programación",
           style = list(fontWeight = "bold",
                        color = "#333")) %>%
  hc_xAxis(title = list(text = NULL)) %>%
  hc_yAxis(title = list(text = "Posición índice TIOBE")) %>%
  hc_yAxis(reversed = TRUE) %>%
  hc_tooltip(pointFormat = "({point.x}, {point.y})") 

df_evolucion

1. Análisis del índice TIOBE

• Mejor lenguaje de cada año
• Nº premios a mejor lenguaje

Código

df_3 <-hall_of_fame()

df_logo <- data.frame(lenguaje = c(2022:2003),
                      logo = c("https://th.bing.com/th?id=OIP.H3I3buZeC8Bkez8ADSrqMwHaHa&w=250&h=250&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.uZuulMpBtfWNUBXf8fHu1AHaHa&w=250&h=250&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.uZuulMpBtfWNUBXf8fHu1AHaHa&w=250&h=250&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.bkbn2-K7c9rMBV5dvYXDrQHaIh&w=233&h=268&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.uZuulMpBtfWNUBXf8fHu1AHaHa&w=250&h=250&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.bkbn2-K7c9rMBV5dvYXDrQHaIh&w=233&h=268&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.7nKR4oRAhgtSbD97mDXWWgHaE8&w=306&h=204&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP._Lm_T3scKhVEVFC54gcRxwHaE8&w=306&h=204&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.Y4VPmWW2m4_V2WFYOEGYRgHaHa&w=250&h=250&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://www.bing.com/th?id=OIP.LN_XZYfNR1okCsPl0pzIqQAAAA&w=114&h=100&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.rI4h3ndENsvdOwunJgW4aQAAAA&w=212&h=212&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.rI4h3ndENsvdOwunJgW4aQAAAA&w=212&h=212&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.uZuulMpBtfWNUBXf8fHu1AHaHa&w=250&h=250&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.7nKR4oRAhgtSbD97mDXWWgHaE8&w=306&h=204&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.bkbn2-K7c9rMBV5dvYXDrQHaIh&w=233&h=268&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.uZuulMpBtfWNUBXf8fHu1AHaHa&w=250&h=250&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.MWdAq1cI3bRY5--dtqIwJwHaHY&w=250&h=249&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP._Lm_T3scKhVEVFC54gcRxwHaE8&w=306&h=204&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.uhYWKsUDBf1rPMjzWZx3OgHaHa&w=250&h=250&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2","https://th.bing.com/th?id=OIP.H3I3buZeC8Bkez8ADSrqMwHaHa&w=250&h=250&c=8&rs=1&qlt=90&o=6&dpr=1.3&pid=3.1&rm=2"))

df_gt <- left_join(df_logo,df_3, by = c("lenguaje" = "Year")) %>%
  rename(Year = lenguaje)
df_gt <- df_gt %>% rename(Año = Year,
                          Ganador = Winner)



tabla_hallfame <- gt(df_gt) %>%
  gtExtras::gt_color_rows( Año,palette = "viridis", pal_type = "continuous") %>%
  gtExtras::gt_img_rows(columns = logo, img_source = 'web') %>%
  tab_header(title = "Mejor lenguaje de programación de cada año",
             subtitle = "Según el índice TIOBE") %>%
  gtExtras::gt_theme_538()

tabla_hallfame

Mejor lenguaje de programación de cada año
Según el índice TIOBE
Año	logo	Ganador
2022		C++
2021		Python
2020		Python
2019		C
2018		Python
2017		C
2016		Go
2015		Java
2014		JavaScript
2013		Transact-SQL
2012		Objective-C
2011		Objective-C
2010		Python
2009		Go
2008		C
2007		Python
2006		Ruby
2005		Java
2004		PHP
2003		C++

Código

df_3 <-hall_of_fame()

df_3 <- df_3 %>% rename( Ganador = `Winner`) %>%
  group_by(Ganador) %>%
  count(n()) %>%
  select(1,3) %>%
  rename(numero_premios = n )


p_winners <- ggplot()+ 
  geom_bar(data = df_3, aes(x = fct_reorder(Ganador, numero_premios), y = numero_premios ,fill = Ganador), stat = "identity") + 
  theme_clean() + 
  labs(title = "Número de premios a Mejor Lenguaje del año",
       caption = "Fuente: Elaboración propia a partir del índice TIOBE",
       y = NULL,
       x = NULL) + 
  theme(legend.position = "none",
        axis.text.y = element_text(size = 12, face = "bold")) +
  scale_fill_brewer(palette = "Paired") + coord_flip()

p_winners

2. Encuesta Stack Overflow

• ¿Con qué lenguajes suelen comenzar los jóvenes?
• Lenguaje más popular por grupo de edad

Código

library(tidyverse)
library(rio)
library(tidyr)
library(here)

my_ruta <- here::here("datos", "survey_results_public.csv")
df <- rio::import(my_ruta)
df <- df %>%  #Nos quedamos solo con dos variables y hacemos una observación por cada fila.
  select(Age , LanguageHaveWorkedWith) %>% 
  separate_rows(LanguageHaveWorkedWith, sep = ";\\s*")

df1 <- df %>% filter(Age == "18-24 years old") %>% 
  filter(!is.na(LanguageHaveWorkedWith)) %>% 
  group_by(LanguageHaveWorkedWith) %>%
  summarise(Observaciones = n()) %>% 
  top_n(10, Observaciones) %>% 
  arrange


colores <- c("#3498db", "#2ecc71", "#e74c3c", "#f39c12", "#9b59b6", "#e67e22", "#1abc9c","#33FF99","#95a5a6", "#FFCCFF")

diez_mas_usados <- ggplot(df1, aes(x = reorder(LanguageHaveWorkedWith, Observaciones), y = Observaciones)) +
  geom_bar(stat = "identity", fill = colores) +
  labs(title = "Top 10 Lenguajes de Programación para Personas de 18-24 años",
       x = "Lenguajes de Programación",
       y = "Número de Observaciones") +
  theme_minimal() +
  theme(axis.title = element_blank(),
        axis.text.x = element_text(size = 8, angle = 45, hjust = 1),
        axis.text.y = element_text(size = 8),
        plot.title = element_text(size = 12, face = "bold"),
        plot.subtitle = element_text(size = 10)) +
  labs(title = "Lenguajes de programación con los que se inician los jóvenes",
       subtitle = "medido en número de respuestas a la encuesta",
       caption = "Fuente: encuesta anual de Stack Overflow") +
  guides(fill = FALSE) + # Eliminar la leyenda de la estética fill
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

diez_mas_usados

Código

#Cargamos librerías
library(gt)
library(gtExtras)
library(RColorBrewer)

df2 <- df %>%
  filter(!is.na(LanguageHaveWorkedWith)) %>%
  group_by(Age, LanguageHaveWorkedWith) %>%
  summarise(Observaciones = n()) %>%
  group_by(Age) %>%
  top_n(1, Observaciones) %>%
  arrange(Age, desc(Observaciones)) %>% 
  select(-Observaciones) %>% 
  rename(lenguaje = LanguageHaveWorkedWith)

df_logos <- data.frame(lenguaje = c("Python", "C", "C++", "Java", "C#", "JavaScript", "PHP", "Visual Basic", "SQL", "Assembly language", "Scratch", "Fortran", "Go", "MATLAB", "Kotlin", "Delphi/Object Pascal", "Swift","Ruby","R","Rust", "Golang"),
                       logo = c("https://tse4.mm.bing.net/th?id=OIP.EDJ9xoErBbZqK2tExVoJfAAAAA&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.bkbn2-K7c9rMBV5dvYXDrQHaIh&pid=Api&rs=1&c=1&qlt=95&w=103&h=118","https://tse1.mm.bing.net/th?id=OIP.ggb-3B-8LRfbgS0lPLYNxwHaIU&pid=Api&rs=1&c=1&qlt=95&w=105&h=118","https://tse1.mm.bing.net/th?id=OIP.dJ7A1hPb-BDRBrhOmhrVAgHaHa&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.2x1yN8k3MVCrkJFR_C5k-wHaIp&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.YPYdVAg5ieRspLI6HIpmDQHaIB&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.pkqphAig1t-PCsy4dkVrfAHaD5&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.Y224uoZnCs8yY0Sos_h4cQHaDd&pid=Api&rs=1&c=1&qlt=95&w=258&h=120","https://tse4.mm.bing.net/th?id=OIP.S_9pgiKacbi3FxPKxDmusgHaFy&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.v5m2HW3FI3RbopnK8Cs6HgAAAA&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.bHt6sdtkwUqP91oPKlruGgHaEM&pid=Api&P=0&h=180","https://tse2.explicit.bing.net/th?id=OIP.tLfIejOZiP_H0Zse0NIn2QAAAA&pid=Api&P=0&h=180","https://tse3.mm.bing.net/th?id=OIP.7nKR4oRAhgtSbD97mDXWWgHaE8&pid=Api&P=0&h=180","https://tse2.mm.bing.net/th?id=OIP.cI8KflTKlnmraGFa22XpKQHaEK&pid=Api&P=0&h=180","https://tse3.mm.bing.net/th?id=OIP.gW_YaXsZmuSntd_sy5zkIwHaHa&pid=Api&P=0&h=180","https://tse2.mm.bing.net/th?id=OIP.hEP7hOlvIzhKU1fGpizd7QAAAA&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.XozeB5BRQSzNCO6k7om62QHaHa&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.2BwE10YiTSXRrnW2LnVh8wHaHa&pid=Api&P=0&h=180","https://tse3.mm.bing.net/th?id=OIP.0rg-o4mEzLO5EHJ1zS52YwHaGe&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.YmPfYoQK3WvGEjFXusKV9QHaHa&pid=Api&P=0&h=180", "https://seeklogo.com/images/G/go-logo-046185B647-seeklogo.com.png"))
df2_final <- left_join(df2, df_logos, by = ("lenguaje"))
#El data.frame daba problemas así que lo he rehecho con un poco de ayuda de R base.
ages <- df2_final$Age
lenguajes <- df2_final$lenguaje
logos <- df2_final$logo
dataframef2_final <- data.frame(ages, lenguajes, logos)

gt_tbl <- dataframef2_final %>%
  gt() %>% 
  gtExtras::gt_theme_espn() %>%
  cols_width("lenguajes" ~ px(100), "logos" ~ px(80)) %>% 
  tab_header(title = "Lenguajes de programación más populares por franjas de edad",
             subtitle = "Según la encuesta anual de Stack Overflow.")%>%
  gtExtras::gt_img_rows(columns = logos, img_source = 'web')

gt_tbl

Lenguajes de programación más populares por franjas de edad
Según la encuesta anual de Stack Overflow.
ages	lenguajes	logos
18-24 years old	JavaScript
25-34 years old	JavaScript
35-44 years old	JavaScript
45-54 years old	JavaScript
55-64 years old	SQL
65 years or older	SQL
Prefer not to say	JavaScript
Under 18 years old	Python

2. Encuesta Stack Overflow

• ¿Cómo se aprende programación?
• Lenguajes de programación con mayores salarios

Código

my_ruta <- here::here("datos", "survey_results_public.csv")
df <- rio::import(my_ruta)
df <- df %>% 
  select(LearnCode) %>% 
  rename(learn=LearnCode) %>% 
  separate_rows(learn, sep = ";\\s*") %>% 
  drop_na() %>% 
  group_by(learn) %>% 
  summarise(Observaciones = n()) %>% 
  arrange(desc(Observaciones)) %>% 
  filter(learn != "Other (please specify):")

colores <- c("#3498db", "#2ecc71", "#e74c3c", "#f39c12", "#9b59b6", "#e67e22", "#1abc9c","#33FF99", "#FFCCFF")
gt_df <- gt(df) %>%
  gtExtras::gt_plt_dot(column = "Observaciones", category_column = `learn`,  
                       max_value = NULL, palette = colores) %>%
  gtExtras::gt_theme_nytimes() %>%
  cols_width(`learn` ~ px(400), 2 ~ px(150)) %>% 
  tab_header(title = "Maneras más usuales que se usan para aprender programación",
             subtitle = "Según la encuesta anual de Stack Overflow.")
gt_df

Maneras más usuales que se usan para aprender programación
Según la encuesta anual de Stack Overflow.
learn	Observaciones
Other online resources (e.g., videos, blogs, forum)	70244
Books / Physical media	45406
School (i.e., University, College, etc)	43957
Online Courses or Certification	43201
On the job training	40380
Colleague	20523
Friend or family member	9936
Coding Bootcamp	8602
Hackathons (virtual or in-person)	7033

Código

my_ruta2 <- here::here("datos","lista_lenguajes_mejor_pagados.xlsx")
aa <- rio::import(my_ruta2)
aa <- aa %>% select(-"Number of Jobs")
df_logos <- data.frame(lenguaje = c("Python", "C", "C++", "Java", "C#", "JavaScript", "PHP", "Visual Basic", "SQL", "Assembly language", "Scratch", "Fortran", "Go", "MATLAB", "Kotlin", "Delphi/Object Pascal", "Swift","Ruby","R","Rust", "Golang"),
                       logo = c("https://tse4.mm.bing.net/th?id=OIP.EDJ9xoErBbZqK2tExVoJfAAAAA&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.bkbn2-K7c9rMBV5dvYXDrQHaIh&pid=Api&rs=1&c=1&qlt=95&w=103&h=118","https://tse1.mm.bing.net/th?id=OIP.ggb-3B-8LRfbgS0lPLYNxwHaIU&pid=Api&rs=1&c=1&qlt=95&w=105&h=118","https://tse1.mm.bing.net/th?id=OIP.dJ7A1hPb-BDRBrhOmhrVAgHaHa&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.2x1yN8k3MVCrkJFR_C5k-wHaIp&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.YPYdVAg5ieRspLI6HIpmDQHaIB&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.pkqphAig1t-PCsy4dkVrfAHaD5&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.Y224uoZnCs8yY0Sos_h4cQHaDd&pid=Api&rs=1&c=1&qlt=95&w=258&h=120","https://tse4.mm.bing.net/th?id=OIP.S_9pgiKacbi3FxPKxDmusgHaFy&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.v5m2HW3FI3RbopnK8Cs6HgAAAA&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.bHt6sdtkwUqP91oPKlruGgHaEM&pid=Api&P=0&h=180","https://tse2.explicit.bing.net/th?id=OIP.tLfIejOZiP_H0Zse0NIn2QAAAA&pid=Api&P=0&h=180","https://tse3.mm.bing.net/th?id=OIP.7nKR4oRAhgtSbD97mDXWWgHaE8&pid=Api&P=0&h=180","https://tse2.mm.bing.net/th?id=OIP.cI8KflTKlnmraGFa22XpKQHaEK&pid=Api&P=0&h=180","https://tse3.mm.bing.net/th?id=OIP.gW_YaXsZmuSntd_sy5zkIwHaHa&pid=Api&P=0&h=180","https://tse2.mm.bing.net/th?id=OIP.hEP7hOlvIzhKU1fGpizd7QAAAA&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.XozeB5BRQSzNCO6k7om62QHaHa&pid=Api&P=0&h=180","https://tse1.mm.bing.net/th?id=OIP.2BwE10YiTSXRrnW2LnVh8wHaHa&pid=Api&P=0&h=180","https://tse3.mm.bing.net/th?id=OIP.0rg-o4mEzLO5EHJ1zS52YwHaGe&pid=Api&P=0&h=180","https://tse4.mm.bing.net/th?id=OIP.YmPfYoQK3WvGEjFXusKV9QHaHa&pid=Api&P=0&h=180", "https://seeklogo.com/images/G/go-logo-046185B647-seeklogo.com.png"))
my_ruta2 <- here::here("datos","lista_lenguajes_mejor_pagados.xlsx")
aa <- rio::import(my_ruta2)
aa <- aa %>% select(-"Number of Jobs") %>% 
  rename(lenguaje = Language)
df_final <- left_join(aa, df_logos, by = ("lenguaje"))

colores <- c("#3498db", "#2ecc71", "#e74c3c", "#f39c12", "#9b59b6", "#e67e22", "#1abc9c","#33FF99","#95a5a6", "#FFCCFF")

gt_salarios <- gt(df_final) %>%
  gtExtras::gt_plt_dot(column = "Median Salary ($)", category_column = `lenguaje`,  
                       max_value = NULL, palette = colores) %>%
  gtExtras::gt_theme_espn() %>%
  cols_width(`lenguaje` ~ px(400), 3 ~ px(80)) %>% 
  tab_header(title = "Lenguajes de programación con mejores salarios",
             subtitle = "Según la web 4 day week.") %>%
  gtExtras::gt_img_rows(columns = logo, img_source = 'web')
gt_salarios

Lenguajes de programación con mejores salarios
Según la web 4 day week.
lenguaje	Median Salary ($)	logo
Swift	116000
Python	105000
Golang	100000
C++	100000
Java	95000
Ruby	95000
Rust	88000
SQL	85000

3. Programación en Economía

• Modelo estocástico neoclásico de crecimiento
• Resultados del paper

Don Jesús “filtro de partículas” Fernández-Villaverde analiza la velocidad de computación de distintos lenguajes, estimando un mismo modelo en los diversos lenguajes¹.

Un planificador social tiene que maximizar un función de valor escogiendo una combinación de consumo \(c_t\) y capital \(k_t\) : \[\max_{\{c_t, k_t+1\}}\mathbb{E}_0\sum_{t=0}^{\infty}(1 - \beta)\beta^t\log c_t\]

Bajo la siguiente restricción presupuestaria: \[c_t + k_{t+1} = z_tk_{t}^\alpha + (1 - \delta)k_t\] El problema a maximizar será: \[V(k,z) = \max_{k'}(1 - \beta)\beta^t\log(zk^\alpha-k')+\beta\mathbb{E}[V(k',z')|z]\]

Código

library(readxl)
library(tidyverse)
library(ggplot2)
library(dplyr)
library(readr)
library(ggthemes)

ex_times <- readxl::read_excel("./datos/execution_times.xlsx") #-importamos la tabla del pdf del paper a una hoja de cálculo, la cargamos, y ya trabajamos con ella en R

ex_times <- ex_times %>%  
  select(1, 4) %>% 
  na.omit() 

colnames(ex_times) <- c("Lenguajes", "times")
ex_times <- ex_times[2:9,] 

ex_times <- ex_times %>% 
  mutate(times = parse_number(times))

ex_times <- ex_times %>%
  mutate(category = case_when(
    as.numeric(times) < 3 ~ "Menos de 3 seg",
    as.numeric(times) >= 3 & as.numeric(times) < 60 ~ "Entre 3 y 60 seg",
    as.numeric(times) >= 60 ~ "Más de 60 seg",
    TRUE ~ NA_character_)) 

ex_times$factor_cat <- as.factor(ex_times$category)
orden_correcto <- c("Menos de 3 seg", "Entre 3 y 60 seg", "Más de 60 seg")
ex_times$factor_cat <- factor(ex_times$category, levels = orden_correcto)

ex_times_plot <- ex_times %>% 
  ggplot() +
  aes(x = Lenguajes, y = times, fill = Lenguajes)  +
  geom_bar(stat = "identity") +
  facet_wrap(~factor_cat, scales = "free") +
  theme_stata(scheme = "s1color") +
  labs(title = "Tiempos de ejecución del modelo", x = "", y = "Tiempo en segundos")

ex_times_plot

1. “Aruoba, S. B., & Fernández-Villaverde, J. (2015). A comparison of programming languages in macroeconomics. Journal of Economic Dynamics and Control, 58, 265-273.”

3.Programación en Economía

• Una miqueta de Econometría
• Gráfico de los residuos
• Autocorrelación

Código

#Modelo econométrico ----------
library(tidyverse)
library(dplyr)
library(reshape2)
library(stargazer)
library(vembedr)
library(jtools)

df <- read.csv("./datos/FebPwtExport12152023.csv")
#- pasamos los datos a formato wide
df <- dcast(df, RegionCode + YearCode ~ VariableCode, value.var = "AggValue")
#- cambiamos el nombre a uno más familiar y creamos la variable trabajo y filtramos los datos a partir del 78
df <- df %>% 
  mutate(l = avh*emp) %>% 
  rename(k = rnna,
         pib = rgdpna,
         year = YearCode) %>% 
  filter(year >= 1960)

modelo <- lm(log(pib) ~ log(l) + log(k), data = df)
modelo
#> 
#> Call:
#> lm(formula = log(pib) ~ log(l) + log(k), data = df)
#> 
#> Coefficients:
#> (Intercept)       log(l)       log(k)  
#>      0.1125       0.2124       0.7473
summ(modelo)

Observations	60
Dependent variable	log(pib)
Type	OLS linear regression

F(2,57)	5635.32
R²	0.99
Adj. R²	0.99

	Est.	S.E.	t val.	p
(Intercept)	0.11	0.37	0.30	0.76
log(l)	0.21	0.04	5.22	0.00
log(k)	0.75	0.01	90.37	0.00
Standard errors: OLS

Código

df_residuos <- data.frame(x = lag(resid(modelo)), y = resid(modelo)) %>% 
  na.omit()

autocorrelacion <- df_residuos %>% 
  ggplot() +
  geom_point(color = "red") +
  aes(x = x, y = y) +
  labs(title = "Dispersión de los residuos del modelo", 
       x = "Residuos rezagados",
       y = "Residuos") +
  theme_clean() + 
  geom_smooth(se = FALSE, method = lm)
autocorrelacion

  

Código

autocorrelación_plot <- df %>% 
  ggplot() + 
  geom_point(color = "#5D8AA8") +
  geom_line(color = "#69b3a2") +
  aes(x = year, y = resid(modelo)) +
  labs(title = "Residuos del modelo", 
       x = "Año",
       y = "Residuos") +
  geom_hline(yintercept = 0, color = "orange", size=.5) +
  theme_classic()
autocorrelación_plot