Ecosistemas
Luis Felipe Franco Rodríguez
2025-08-20
# Tema 3: Modelos de Ecuaciones Estructurales
###Calidad del Suelo:
SPH: pH del Suelo
NC: Contenido de Nutrientes
OM: Materia Orgánica
###Calidad del Agua:
CL: Niveles de Contaminantes
DO: Oxígeno Disuelto
WPH: pH del Agua
###Salud del Ecosistema
SD: Diversidad de Especies
BM: Biomasa
EP: Productividad del Ecosistema
#Cargar Base de datos y librerias
library(tidyverse)## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.4 ✔ readr 2.1.5
## ✔ forcats 1.0.0 ✔ stringr 1.5.1
## ✔ ggplot2 3.5.2 ✔ tibble 3.3.0
## ✔ lubridate 1.9.4 ✔ tidyr 1.3.1
## ✔ purrr 1.1.0
## ── 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 errors# install.packages("gplots")
library(gplots)##
## Adjuntando el paquete: 'gplots'
##
## The following object is masked from 'package:stats':
##
## lowess# install.packages("plm")
library(plm)##
## Adjuntando el paquete: 'plm'
##
## The following objects are masked from 'package:dplyr':
##
## between, lag, lead# install.packages("DataExplorer")
library(DataExplorer)
# install.packages("forecast")
library(forecast)## Registered S3 method overwritten by 'quantmod':
## method from
## as.zoo.data.frame zoo# install.packages("lavaan")
library(lavaan)## This is lavaan 0.6-19
## lavaan is FREE software! Please report any bugs.# install.packages("lavaanPlot")
library(lavaanPlot)library(readr)
df <- read.csv("~/ecosistema.csv")modelo <- '
# Variables latentes
Calidad_del_Suelo =~ SPH + NC + OM
Calidad_del_Agua =~ CL + DO + WPH
Salud_del_Ecosistema =~ SD + BM + EP
# Covarianzas
Calidad_del_Suelo ~~ Calidad_del_Agua
Salud_del_Ecosistema ~~ Calidad_del_Agua
Salud_del_Ecosistema ~~ Calidad_del_Suelo
'## Generar el diagrama
df3 <- scale(df)
df4 <- cfa(modelo, df3)## Warning: lavaan->lav_object_post_check():
## covariance matrix of latent variables is not positive definite ; use
## lavInspect(fit, "cov.lv") to investigate.summary(df4)## lavaan 0.6-19 ended normally after 97 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 21
##
## Number of observations 200
##
## Model Test User Model:
##
## Test statistic 17.149
## Degrees of freedom 24
## P-value (Chi-square) 0.842
##
## Parameter Estimates:
##
## Standard errors Standard
## Information Expected
## Information saturated (h1) model Structured
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## Calidad_del_Suelo =~
## SPH 1.000
## NC 2.217 1.332 1.664 0.096
## OM 0.167 0.402 0.414 0.679
## Calidad_del_Agua =~
## CL 1.000
## DO -0.827 0.427 -1.936 0.053
## WPH 0.404 0.359 1.124 0.261
## Salud_del_Ecosistema =~
## SD 1.000
## BM -1.899 3.995 -0.475 0.634
## EP -4.224 8.093 -0.522 0.602
##
## Covariances:
## Estimate Std.Err z-value P(>|z|)
## Calidad_del_Suelo ~~
## Calidad_del_Ag -0.079 0.051 -1.558 0.119
## Calidad_del_Agua ~~
## Sald_dl_Ecsstm -0.018 0.034 -0.513 0.608
## Calidad_del_Suelo ~~
## Sald_dl_Ecsstm -0.010 0.020 -0.497 0.619
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .SPH 0.926 0.103 9.016 0.000
## .NC 0.657 0.226 2.906 0.004
## .OM 0.993 0.100 9.978 0.000
## .CL 0.953 0.118 8.086 0.000
## .DO 0.966 0.108 8.977 0.000
## .WPH 0.988 0.099 9.938 0.000
## .SD 0.992 0.100 9.954 0.000
## .BM 0.985 0.102 9.618 0.000
## .EP 0.948 0.166 5.720 0.000
## Calidad_del_Sl 0.069 0.057 1.197 0.231
## Calidad_del_Ag 0.042 0.075 0.558 0.577
## Sald_dl_Ecsstm 0.003 0.012 0.221 0.825lavaanPlot(df4, coef = TRUE, cov = TRUE)# Actividad Hogares
library(readxl)
df1 <- read_excel("~/hogares.xlsx")
summary(df1)## HogarID Año Miembros Ingreso Gasto
## Min. : 1.00 Min. :2010 Min. :2.00 Min. : 5039 Min. : 3273
## 1st Qu.: 25.75 1st Qu.:2012 1st Qu.:3.00 1st Qu.:21198 1st Qu.:14474
## Median : 50.50 Median :2014 Median :4.00 Median :26854 Median :20135
## Mean : 50.50 Mean :2014 Mean :3.52 Mean :27066 Mean :19879
## 3rd Qu.: 75.25 3rd Qu.:2017 3rd Qu.:4.25 3rd Qu.:32627 3rd Qu.:25113
## Max. :100.00 Max. :2019 Max. :5.00 Max. :50129 Max. :34753
## Ahorro Satisfacción
## Min. :-23306.6 Min. : 1.157
## 1st Qu.: 304.7 1st Qu.: 4.796
## Median : 6608.9 Median : 5.667
## Mean : 7186.3 Mean : 5.697
## 3rd Qu.: 13555.9 3rd Qu.: 6.607
## Max. : 38655.5 Max. :10.000# Generar conjunto de datos de panel (índices correctos)
panel_1 <- pdata.frame(df1, index = c("HogarID","Año"))
head(panel_1)## HogarID Año Miembros Ingreso Gasto Ahorro Satisfacción
## 1-2010 1 2010 2 41372.72 23948.71 17424.005 5.544594
## 1-2011 1 2011 2 49362.29 25511.14 23851.155 6.456287
## 1-2012 1 2012 2 34227.85 24558.43 9669.414 5.484323
## 1-2013 1 2013 2 39474.82 26494.29 12980.538 6.416048
## 1-2014 1 2014 2 45965.89 30411.14 15554.746 4.116369
## 1-2015 1 2015 2 41671.21 25041.87 16629.338 6.499067## Prueba de heterogenidad
plotmeans(Ingreso ~ HogarID,
main = "Prueba de Heterogeneidad entre hogares para el ingreso",
data = panel_1)
## Modelos de panel con las variables de
hogares
# Modelo 1. Regresión agrupada (pooled)
pooled <- plm(Ingreso ~ Satisfacción, data = panel_1, model = "pooling")
summary(pooled)## Pooling Model
##
## Call:
## plm(formula = Ingreso ~ Satisfacción, data = panel_1, model = "pooling")
##
## Balanced Panel: n = 100, T = 10, N = 1000
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -20196.53 -5106.46 -575.98 5095.02 23468.66
##
## Coefficients:
## Estimate Std. Error t-value Pr(>|t|)
## (Intercept) 10597.75 976.80 10.850 < 2.2e-16 ***
## Satisfacción 2890.77 166.68 17.343 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 6.8145e+10
## Residual Sum of Squares: 5.2364e+10
## R-Squared: 0.23158
## Adj. R-Squared: 0.23081
## F-statistic: 300.772 on 1 and 998 DF, p-value: < 2.22e-16# Modelo 2. Efectos Fijos (within)
within <- plm(Ingreso ~ Satisfacción, data = panel_1, model = "within")
summary(within)## Oneway (individual) effect Within Model
##
## Call:
## plm(formula = Ingreso ~ Satisfacción, data = panel_1, model = "within")
##
## Balanced Panel: n = 100, T = 10, N = 1000
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -15591.951 -3123.123 -74.284 3010.168 13134.979
##
## Coefficients:
## Estimate Std. Error t-value Pr(>|t|)
## Satisfacción 1698.14 132.73 12.794 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 2.3013e+10
## Residual Sum of Squares: 1.9469e+10
## R-Squared: 0.15403
## Adj. R-Squared: 0.05993
## F-statistic: 163.687 on 1 and 899 DF, p-value: < 2.22e-16# Prueba F (pooled vs within)
pFtest(within, pooled)##
## F test for individual effects
##
## data: Ingreso ~ Satisfacción
## F = 15.343, df1 = 99, df2 = 899, p-value < 2.2e-16
## alternative hypothesis: significant effects# Modelos de efectos aleatorios
# Efectos aleatorios
walhus <- plm(Ingreso ~ Satisfacción, data = panel_1, model = "random", random.method = "walhus")
amemiya <- plm(Ingreso ~ Satisfacción, data = panel_1, model = "random", random.method = "amemiya")
nerlove <- plm(Ingreso ~ Satisfacción, data = panel_1, model = "random", random.method = "nerlove")
summary(walhus); summary(amemiya); summary(nerlove)## Oneway (individual) effect Random Effect Model
## (Wallace-Hussain's transformation)
##
## Call:
## plm(formula = Ingreso ~ Satisfacción, data = panel_1, model = "random",
## random.method = "walhus")
##
## Balanced Panel: n = 100, T = 10, N = 1000
##
## Effects:
## var std.dev share
## idiosyncratic 23574420 4855 0.45
## individual 28789336 5366 0.55
## theta: 0.7249
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -16507.33 -3220.23 -147.96 3184.91 15215.46
##
## Coefficients:
## Estimate Std. Error z-value Pr(>|z|)
## (Intercept) 16632.69 925.15 17.978 < 2.2e-16 ***
## Satisfacción 1831.41 131.69 13.907 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 2.6429e+10
## Residual Sum of Squares: 2.2139e+10
## R-Squared: 0.16233
## Adj. R-Squared: 0.16149
## Chisq: 193.404 on 1 DF, p-value: < 2.22e-16## Oneway (individual) effect Random Effect Model
## (Amemiya's transformation)
##
## Call:
## plm(formula = Ingreso ~ Satisfacción, data = panel_1, model = "random",
## random.method = "amemiya")
##
## Balanced Panel: n = 100, T = 10, N = 1000
##
## Effects:
## var std.dev share
## idiosyncratic 21631698 4651 0.393
## individual 33418160 5781 0.607
## theta: 0.7534
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -16370.54 -3188.47 -210.78 3188.52 14905.18
##
## Coefficients:
## Estimate Std. Error z-value Pr(>|z|)
## (Intercept) 16777.35 953.98 17.587 < 2.2e-16 ***
## Satisfacción 1806.01 130.63 13.825 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 2.5757e+10
## Residual Sum of Squares: 2.1617e+10
## R-Squared: 0.16074
## Adj. R-Squared: 0.1599
## Chisq: 191.14 on 1 DF, p-value: < 2.22e-16## Oneway (individual) effect Random Effect Model
## (Nerlove's transformation)
##
## Call:
## plm(formula = Ingreso ~ Satisfacción, data = panel_1, model = "random",
## random.method = "nerlove")
##
## Balanced Panel: n = 100, T = 10, N = 1000
##
## Effects:
## var std.dev share
## idiosyncratic 19468528 4412 0.351
## individual 35940737 5995 0.649
## theta: 0.7733
##
## Residuals:
## Min. 1st Qu. Median 3rd Qu. Max.
## -16275.51 -3113.76 -212.49 3188.29 14690.19
##
## Coefficients:
## Estimate Std. Error z-value Pr(>|z|)
## (Intercept) 16869.92 981.37 17.190 < 2.2e-16 ***
## Satisfacción 1789.76 129.95 13.773 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Total Sum of Squares: 2.5332e+10
## Residual Sum of Squares: 2.1286e+10
## R-Squared: 0.15972
## Adj. R-Squared: 0.15888
## Chisq: 189.701 on 1 DF, p-value: < 2.22e-16# Hausman (fixed vs random). Usar 'swar' suele ser estable para la prueba:
random_swar <- plm(Ingreso ~ Satisfacción, data = panel_1, model = "random", random.method = "swar")
phtest(within, random_swar)##
## Hausman Test
##
## data: Ingreso ~ Satisfacción
## chisq = 67.193, df = 1, p-value = 2.462e-16
## alternative hypothesis: one model is inconsistent# Estados de Mexico
# devtools::install_github("diegovalle/mxmaps")
# library(mxmaps)
# df2 <- df_mxstate_2020
# df_mxstate_2020$value <- df2$pop # reemplaza con tus valores
# mxstate_choropleth(df_mxstate_2020)# Importar base de datos
df3 <- read.csv("~/population.csv")df2 <- read.csv("~/population.csv")
df3 <- df2 %>% dplyr::filter(state == "TX")
ts_tx <- ts(df3$population, start = 1900, frequency = 1)
arima_tx <- auto.arima(ts_tx)
pronostico <- forecast(arima_tx, level = c(95), h = 31)
pronostico## Point Forecast Lo 95 Hi 95
## 2020 29398472 29199487 29597457
## 2021 29806827 29463665 30149990
## 2022 30215183 29742956 30687410
## 2023 30623538 30024100 31222977
## 2024 31031894 30303359 31760429
## 2025 31440249 30579246 32301253
## 2026 31848605 30851090 32846119
## 2027 32256960 31118581 33395339
## 2028 32665316 31381587 33949044
## 2029 33073671 31640070 34507272
## 2030 33482027 31894047 35070007
## 2031 33890382 32143561 35637204
## 2032 34298738 32388674 36208801
## 2033 34707093 32629456 36784730
## 2034 35115449 32865983 37364914
## 2035 35523804 33098330 37949278
## 2036 35932160 33326573 38537746
## 2037 36340515 33550788 39130242
## 2038 36748871 33771046 39726695
## 2039 37157226 33987418 40327034
## 2040 37565581 34199972 40931191
## 2041 37973937 34408774 41539100
## 2042 38382292 34613887 42150698
## 2043 38790648 34815371 42765925
## 2044 39199003 35013284 43384723
## 2045 39607359 35207682 44007036
## 2046 40015714 35398618 44632810
## 2047 40424070 35586145 45261995
## 2048 40832425 35770311 45894540
## 2049 41240781 35951163 46530399
## 2050 41649136 36128748 47169524plot(pronostico, main = "Población en Texas")