Summary global
NMR
2024-10-02
Resultados globales medidas precisión y exactitud cada método
library(readxl)
SUMMARY <- read_excel("SUMMARY.xlsx")
df_global <- data.frame(
Metodo = c("PA", "MAP", "EGA", "KP_mc", "KP_mcm", "KP"),
mPC = numeric(6),
mMBE = numeric(6),
mMAE = numeric(6)
)
df_global[1, 2] <- mean(c(SUMMARY$PC_pasp, SUMMARY$PC_pasp))
df_global[2, 2] <- mean(c(SUMMARY$PC_mapp, SUMMARY$PC_mapsp))
df_global[3, 2] <- mean(c(SUMMARY$PC_egap, SUMMARY$PC_egasp))
df_global[4, 2] <- mean(c(SUMMARY$PC_kpp, SUMMARY$PC_kpsp))
df_global[5, 2] <- mean(c(SUMMARY$PC_kppm, SUMMARY$PC_kpspm))
df_global[6, 2] <- mean(c(SUMMARY$PC_kpd))
df_global[1, 3] <- mean(c(SUMMARY$MBE_pasp, SUMMARY$MBE_pasp))
df_global[2, 3] <- mean(c(SUMMARY$MBE_mapp, SUMMARY$MBE_mapsp))
df_global[3, 3] <- mean(c(SUMMARY$MBE_egap, SUMMARY$MBE_egasp))
df_global[4, 3] <- mean(c(SUMMARY$MBE_kpp, SUMMARY$MBE_kpsp))
df_global[5, 3] <- mean(c(SUMMARY$MBE_kppm, SUMMARY$MBE_kpspm))
df_global[6, 3] <- mean(c(SUMMARY$MBE_kpd))
df_global[1, 4] <- mean(c(SUMMARY$MAE_pasp, SUMMARY$MAE_pasp))
df_global[2, 4] <- mean(c(SUMMARY$MAE_mapp, SUMMARY$MAE_mapsp))
df_global[3, 4] <- mean(c(SUMMARY$MAE_egap, SUMMARY$MAE_egasp))
df_global[4, 4] <- mean(c(SUMMARY$MAE_kpp, SUMMARY$MAE_kpsp))
df_global[5, 4] <- mean(c(SUMMARY$MAE_kppm, SUMMARY$MAE_kpspm))
df_global[6, 4] <- mean(c(SUMMARY$MAE_kpd))
library(knitr)
library(kableExtra)
kable(df_global, format = "html", table.attr = "class='table table-striped'") %>%
kable_styling(full_width = FALSE)| Metodo | mPC | mMBE | mMAE |
|---|---|---|---|
| PA | 0.1895250 | -2.3405292 | 2.405379 |
| MAP | 0.1766250 | 2.2611521 | 2.288340 |
| EGA | 0.4800833 | 0.5948854 | 1.215235 |
| KP_mc | 0.3176354 | -0.4221125 | 1.208187 |
| KP_mcm | 0.0527167 | -3.3981375 | 3.417075 |
| KP | 0.1317542 | -3.7780458 | 3.779104 |
*PA = Análisis paralelo, MAP = Mínimo promedio parcial, EGA = Análisis exploratorio gráfico, KP_mc = Kaiser-Princals usando la matriz de correlación, KP_mcm = Kaiser princals usando la matriz de correlación ajustando el nivel métrico, KP = Kaiser Princals aplicado a la base de datos de tipo ordinal
library(dplyr)##
## Adjuntando el paquete: 'dplyr'## The following object is masked from 'package:kableExtra':
##
## group_rows## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionHFL_O <- SUMMARY %>% filter(Loading == "HighFL_O")
LFL_O <- SUMMARY %>% filter(Loading == "LowFL_O")
HFL_03 <- SUMMARY %>% filter(Loading == "HighFL_cor03")
LFL_03 <- SUMMARY %>% filter(Loading == "LowFL_cor03")
HFL_05 <- SUMMARY %>% filter(Loading == "HighFL_cor05")
LFL_05 <- SUMMARY %>% filter(Loading == "LowFL_cor05")
df_load_corf <- data.frame(
Metodo = c(rep("PA",6), rep("MAP",6), rep("EGA", 6), rep("KP_mc",6), rep("KP_mcm", 6), rep("KP",6)),
Carga_corf = c("HFL_O", "LFL_O", "HFL_03", "LFL_03", "HFL_05", "LFL_05", "HFL_O", "LFL_O", "HFL_03", "LFL_03", "HFL_05", "LFL_05", "HFL_O", "LFL_O", "HFL_03", "LFL_03", "HFL_05", "LFL_05", "HFL_O", "LFL_O", "HFL_03", "LFL_03", "HFL_05", "LFL_05", "HFL_O", "LFL_O", "HFL_03", "LFL_03", "HFL_05", "LFL_05", "HFL_O", "LFL_O", "HFL_03", "LFL_03", "HFL_05", "LFL_05"),
mPC = numeric(36),
mMBE = numeric(36),
mMAE = numeric(36)
)
df_load_corf[1, 3] <- mean(c(HFL_O$PC_pasp, HFL_O$PC_pasp))
df_load_corf[2, 3] <- mean(c(LFL_O$PC_pasp, HFL_O$PC_pasp))
df_load_corf[3, 3] <- mean(c(HFL_03$PC_pasp, HFL_03$PC_pasp))
df_load_corf[4, 3] <- mean(c(LFL_03$PC_pasp, LFL_03$PC_pasp))
df_load_corf[5, 3] <- mean(c(HFL_05$PC_pasp, HFL_05$PC_pasp))
df_load_corf[6, 3] <- mean(c(LFL_05$PC_pasp, LFL_05$PC_pasp))
df_load_corf[7, 3] <- mean(c(HFL_O$PC_mapp, HFL_O$PC_mapsp))
df_load_corf[8, 3] <- mean(c(LFL_O$PC_mapp, LFL_O$PC_mapsp))
df_load_corf[9, 3] <- mean(c(HFL_03$PC_mapp, HFL_03$PC_mapsp))
df_load_corf[10, 3] <- mean(c(LFL_03$PC_mapp, LFL_03$PC_mapsp))
df_load_corf[11, 3] <- mean(c(HFL_05$PC_mapp, HFL_05$PC_mapsp))
df_load_corf[12, 3] <- mean(c(LFL_05$PC_mapp, LFL_05$PC_mapsp))
df_load_corf[13, 3] <- mean(c(HFL_O$PC_egap, HFL_O$PC_egasp))
df_load_corf[14, 3] <- mean(c(LFL_O$PC_egap, LFL_O$PC_egasp))
df_load_corf[15, 3] <- mean(c(HFL_03$PC_egap, HFL_03$PC_egasp))
df_load_corf[16, 3] <- mean(c(LFL_03$PC_egap, LFL_03$PC_egasp))
df_load_corf[17, 3] <- mean(c(HFL_05$PC_egap, HFL_05$PC_egasp))
df_load_corf[18, 3] <- mean(c(LFL_05$PC_egap, LFL_05$PC_egasp))
df_load_corf[19, 3] <- mean(c(HFL_O$PC_kpp, HFL_O$PC_kpsp))
df_load_corf[20, 3] <- mean(c(LFL_O$PC_kpp, LFL_O$PC_kpsp))
df_load_corf[21, 3] <- mean(c(HFL_03$PC_kpp, HFL_03$PC_kpsp))
df_load_corf[22, 3] <- mean(c(LFL_03$PC_kpp, LFL_03$PC_kpsp))
df_load_corf[23, 3] <- mean(c(HFL_05$PC_kpp, HFL_05$PC_kpsp))
df_load_corf[24, 3] <- mean(c(LFL_05$PC_kpp, LFL_05$PC_kpsp))
df_load_corf[25, 3] <- mean(c(HFL_O$PC_kppm, HFL_O$PC_kpspm))
df_load_corf[26, 3] <- mean(c(LFL_O$PC_kppm, LFL_O$PC_kpspm))
df_load_corf[27, 3] <- mean(c(HFL_03$PC_kppm, HFL_03$PC_kpspm))
df_load_corf[28, 3] <- mean(c(LFL_03$PC_kppm, LFL_03$PC_kpspm))
df_load_corf[29, 3] <- mean(c(HFL_05$PC_kppm, HFL_05$PC_kpspm))
df_load_corf[30, 3] <- mean(c(LFL_05$PC_kpd))
df_load_corf[31, 3] <- mean(c(HFL_O$PC_kpd))
df_load_corf[32, 3] <- mean(c(LFL_O$PC_kpd))
df_load_corf[33, 3] <- mean(c(HFL_03$PC_kpd))
df_load_corf[34, 3] <- mean(c(LFL_03$PC_kpd))
df_load_corf[35, 3] <- mean(c(HFL_05$PC_kpd))
df_load_corf[36, 3] <- mean(c(LFL_05$PC_kpd))
df_load_corf[1, 4] <- mean(c(HFL_O$MBE_pasp, HFL_O$MBE_pasp))
df_load_corf[2, 4] <- mean(c(LFL_O$MBE_pasp, HFL_O$MBE_pasp))
df_load_corf[3, 4] <- mean(c(HFL_03$MBE_pasp, HFL_03$MBE_pasp))
df_load_corf[4, 4] <- mean(c(LFL_03$MBE_pasp, LFL_03$MBE_pasp))
df_load_corf[5, 4] <- mean(c(HFL_05$MBE_pasp, HFL_05$MBE_pasp))
df_load_corf[6, 4] <- mean(c(LFL_05$MBE_pasp, LFL_05$MBE_pasp))
df_load_corf[7, 4] <- mean(c(HFL_O$MBE_mapp, HFL_O$MBE_mapsp))
df_load_corf[8, 4] <- mean(c(LFL_O$MBE_mapp, LFL_O$MBE_mapsp))
df_load_corf[9, 4] <- mean(c(HFL_03$MBE_mapp, HFL_03$MBE_mapsp))
df_load_corf[10, 4] <- mean(c(LFL_03$MBE_mapp, LFL_03$MBE_mapsp))
df_load_corf[11, 4] <- mean(c(HFL_05$MBE_mapp, HFL_05$MBE_mapsp))
df_load_corf[12, 4] <- mean(c(LFL_05$MBE_mapp, LFL_05$MBE_mapsp))
df_load_corf[13, 4] <- mean(c(HFL_O$MBE_egap, HFL_O$MBE_egasp))
df_load_corf[14, 4] <- mean(c(LFL_O$MBE_egap, LFL_O$MBE_egasp))
df_load_corf[15, 4] <- mean(c(HFL_03$MBE_egap, HFL_03$MBE_egasp))
df_load_corf[16, 4] <- mean(c(LFL_03$MBE_egap, LFL_03$MBE_egasp))
df_load_corf[17, 4] <- mean(c(HFL_05$MBE_egap, HFL_05$MBE_egasp))
df_load_corf[18, 4] <- mean(c(LFL_05$MBE_egap, LFL_05$MBE_egasp))
df_load_corf[19, 4] <- mean(c(HFL_O$MBE_kpp, HFL_O$MBE_kpsp))
df_load_corf[20, 4] <- mean(c(LFL_O$MBE_kpp, LFL_O$MBE_kpsp))
df_load_corf[21, 4] <- mean(c(HFL_03$MBE_kpp, HFL_03$MBE_kpsp))
df_load_corf[22, 4] <- mean(c(LFL_03$MBE_kpp, LFL_03$MBE_kpsp))
df_load_corf[23, 4] <- mean(c(HFL_05$MBE_kpp, HFL_05$MBE_kpsp))
df_load_corf[24, 4] <- mean(c(LFL_05$MBE_kpp, LFL_05$MBE_kpsp))
df_load_corf[25, 4] <- mean(c(HFL_O$MBE_kppm, HFL_O$MBE_kpspm))
df_load_corf[26, 4] <- mean(c(LFL_O$MBE_kppm, LFL_O$MBE_kpspm))
df_load_corf[27, 4] <- mean(c(HFL_03$MBE_kppm, HFL_03$MBE_kpspm))
df_load_corf[28, 4] <- mean(c(LFL_03$MBE_kppm, LFL_03$MBE_kpspm))
df_load_corf[29, 4] <- mean(c(HFL_05$MBE_kppm, HFL_05$MBE_kpspm))
df_load_corf[30, 4] <- mean(c(LFL_05$MBE_kpd))
df_load_corf[31, 4] <- mean(c(HFL_O$MBE_kpd))
df_load_corf[32, 4] <- mean(c(LFL_O$MBE_kpd))
df_load_corf[33, 4] <- mean(c(HFL_03$MBE_kpd))
df_load_corf[34, 4] <- mean(c(LFL_03$MBE_kpd))
df_load_corf[35, 4] <- mean(c(HFL_05$MBE_kpd))
df_load_corf[36, 4] <- mean(c(LFL_05$MBE_kpd))
df_load_corf[1, 5] <- mean(c(HFL_O$MAE_pasp, HFL_O$MAE_pasp))
df_load_corf[2, 5] <- mean(c(LFL_O$MAE_pasp, HFL_O$MAE_pasp))
df_load_corf[3, 5] <- mean(c(HFL_03$MAE_pasp, HFL_03$MAE_pasp))
df_load_corf[4, 5] <- mean(c(LFL_03$MAE_pasp, LFL_03$MAE_pasp))
df_load_corf[5, 5] <- mean(c(HFL_05$MAE_pasp, HFL_05$MAE_pasp))
df_load_corf[6, 5] <- mean(c(LFL_05$MAE_pasp, LFL_05$MAE_pasp))
df_load_corf[7, 5] <- mean(c(HFL_O$MAE_mapp, HFL_O$MAE_mapsp))
df_load_corf[8, 5] <- mean(c(LFL_O$MAE_mapp, LFL_O$MAE_mapsp))
df_load_corf[9, 5] <- mean(c(HFL_03$MAE_mapp, HFL_03$MAE_mapsp))
df_load_corf[10, 5] <- mean(c(LFL_03$MAE_mapp, LFL_03$MAE_mapsp))
df_load_corf[11, 5] <- mean(c(HFL_05$MAE_mapp, HFL_05$MAE_mapsp))
df_load_corf[12, 5] <- mean(c(LFL_05$MAE_mapp, LFL_05$MAE_mapsp))
df_load_corf[13, 5] <- mean(c(HFL_O$MAE_egap, HFL_O$MAE_egasp))
df_load_corf[14, 5] <- mean(c(LFL_O$MAE_egap, LFL_O$MAE_egasp))
df_load_corf[15, 5] <- mean(c(HFL_03$MAE_egap, HFL_03$MAE_egasp))
df_load_corf[16, 5] <- mean(c(LFL_03$MAE_egap, LFL_03$MAE_egasp))
df_load_corf[17, 5] <- mean(c(HFL_05$MAE_egap, HFL_05$MAE_egasp))
df_load_corf[18, 5] <- mean(c(LFL_05$MAE_egap, LFL_05$MAE_egasp))
df_load_corf[19, 5] <- mean(c(HFL_O$MAE_kpp, HFL_O$MAE_kpsp))
df_load_corf[20, 5] <- mean(c(LFL_O$MAE_kpp, LFL_O$MAE_kpsp))
df_load_corf[21, 5] <- mean(c(HFL_03$MAE_kpp, HFL_03$MAE_kpsp))
df_load_corf[22, 5] <- mean(c(LFL_03$MAE_kpp, LFL_03$MAE_kpsp))
df_load_corf[23, 5] <- mean(c(HFL_05$MAE_kpp, HFL_05$MAE_kpsp))
df_load_corf[24, 5] <- mean(c(LFL_05$MAE_kpp, LFL_05$MAE_kpsp))
df_load_corf[25, 5] <- mean(c(HFL_O$MAE_kppm, HFL_O$MAE_kpspm))
df_load_corf[26, 5] <- mean(c(LFL_O$MAE_kppm, LFL_O$MAE_kpspm))
df_load_corf[27, 5] <- mean(c(HFL_03$MAE_kppm, HFL_03$MAE_kpspm))
df_load_corf[28, 5] <- mean(c(LFL_03$MAE_kppm, LFL_03$MAE_kpspm))
df_load_corf[29, 5] <- mean(c(HFL_05$MAE_kppm, HFL_05$MAE_kpspm))
df_load_corf[30, 5] <- mean(c(LFL_05$MAE_kpd))
df_load_corf[31, 5] <- mean(c(HFL_O$MAE_kpd))
df_load_corf[32, 5] <- mean(c(LFL_O$MAE_kpd))
df_load_corf[33, 5] <- mean(c(HFL_03$MAE_kpd))
df_load_corf[34, 5] <- mean(c(LFL_03$MAE_kpd))
df_load_corf[35, 5] <- mean(c(HFL_05$MAE_kpd))
df_load_corf[36, 5] <- mean(c(LFL_05$MAE_kpd))
library(knitr)
library(kableExtra)
kable(df_load_corf, format = "html", table.attr = "class='table table-striped'") %>%
kable_styling(full_width = FALSE)| Metodo | Carga_corf | mPC | mMBE | mMAE |
|---|---|---|---|---|
| PA | HFL_O | 0.3450556 | -1.2884722 | 1.3045278 |
| PA | LFL_O | 0.2101250 | -2.3113333 | 2.3416944 |
| PA | HFL_03 | 0.2895833 | -1.4966111 | 1.5391111 |
| PA | LFL_03 | 0.0722222 | -3.4321667 | 3.4885556 |
| PA | HFL_05 | 0.2154722 | -1.6869722 | 1.8191944 |
| PA | LFL_05 | 0.0767222 | -3.5700000 | 3.6507222 |
| MAP | HFL_O | 0.4107361 | 1.8328472 | 1.9126528 |
| MAP | LFL_O | 0.0335556 | 3.0614583 | 3.0715972 |
| MAP | HFL_03 | 0.2948056 | 1.7496528 | 1.8045417 |
| MAP | LFL_03 | 0.0198611 | 2.9944306 | 3.0008750 |
| MAP | HFL_05 | 0.1303056 | 2.0650556 | 2.0892778 |
| MAP | LFL_05 | 0.0098194 | 2.9826528 | 2.9884028 |
| EGA | HFL_O | 0.7505694 | 0.0686250 | 0.4905139 |
| EGA | LFL_O | 0.2785972 | 1.0758333 | 1.8803611 |
| EGA | HFL_03 | 0.6765278 | 0.1401806 | 0.6070417 |
| EGA | LFL_03 | 0.2404306 | 1.1583056 | 2.0021389 |
| EGA | HFL_05 | 0.5379028 | 0.3485278 | 0.8345556 |
| EGA | LFL_05 | 0.2096944 | 1.3055694 | 2.1039306 |
| KP_mc | HFL_O | 0.3499444 | 0.3869722 | 0.9120278 |
| KP_mc | LFL_O | 0.3520417 | -0.5445556 | 1.2710278 |
| KP_mc | HFL_03 | 0.3671250 | 0.0219444 | 1.0165556 |
| KP_mc | LFL_03 | 0.3090833 | -0.8846806 | 1.5223750 |
| KP_mc | HFL_05 | 0.3364167 | -0.2413056 | 1.1463056 |
| KP_mc | LFL_05 | 0.2780694 | -1.0081111 | 1.6419444 |
| KP_mcm | HFL_O | 0.2266944 | -1.7534167 | 1.8705556 |
| KP_mcm | LFL_O | 0.0132500 | -3.8801111 | 3.8805556 |
| KP_mcm | HFL_03 | 0.0764306 | -2.9044444 | 2.9120556 |
| KP_mcm | LFL_03 | 0.0046528 | -4.3981250 | 4.3982361 |
| KP_mcm | HFL_05 | 0.0229444 | -3.6120694 | 3.6130139 |
| KP_mcm | LFL_05 | 0.0533333 | -4.6880000 | 4.6880000 |
| KP | HFL_O | 0.1535556 | -3.5571944 | 3.5571944 |
| KP | LFL_O | 0.0397778 | -5.0394444 | 5.0394444 |
| KP | HFL_03 | 0.1702500 | -3.3644167 | 3.3644722 |
| KP | LFL_03 | 0.0447778 | -4.8764444 | 4.8764444 |
| KP | HFL_05 | 0.2055000 | -3.1456667 | 3.1526667 |
| KP | LFL_05 | 0.0533333 | -4.6880000 | 4.6880000 |
library(ggplot2)
df_load_corf$Metodo <- as.factor(df_load_corf$Metodo)
df_load_corf$Carga_corf <- as.factor(df_load_corf$Carga_corf)
ggplot(df_load_corf, aes(x = Carga_corf, y = mMBE, color = Metodo)) +
geom_point(position = position_dodge(width = 0.5)) +
geom_errorbar(aes(ymin = mMBE - mMAE, ymax = mMBE + mMAE),
width = 0.2, position = position_dodge(width = 0.5)) +
labs(title = "Desviación de MBE con MAE como Error - carga factorial - cor entre factores",
x = "Carga factorial - relacion entre factores",
y = "MBE (con MAE como barras de error)") +
theme_minimal()
# Correcta proporcion
ggplot(df_load_corf, aes(x = Carga_corf, y = mPC, fill = Metodo)) +
geom_bar(stat = "identity", position = "dodge") +
labs(title = "Proporciones Correctas por carga factorial y relacion entre factores",
x = "Condición",
y = "Proporción Correcta") +
scale_fill_grey(start = 0, end = 0.8) +
theme_minimal()
library(tidyr)
df_long <- df_load_corf%>%
pivot_longer(cols = c(mPC, mMBE, mMAE),
names_to = "variable",
values_to = "value")
ggplot(df_long, aes(x = Carga_corf, y = value, group = variable, color = variable)) +
geom_line() +
geom_point() +
facet_wrap(~ Metodo, scales = "free_y") + # Facetas según el método
labs(title = "Gráfico Línea para PC, MBE, y MAE según Método",
x = "Carga Factorial - CORRELACION ENTRE FACTORES",
y = "Valor") +
theme_minimal() +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) 
Resultado por método y tipo de correlación empleado
df_glob_cor <- data.frame(
Metodo = c("PAp","Pasp" ,"MAPp", "MAPsp", "EGAp","EGAsp", "KP_mcp", "KP_mcsp",
"KP_mcmp", "KP_mcmsp","KP"),
mPC = numeric(11),
mMBE = numeric(11),
mMAE = numeric(11)
)
df_glob_cor[1, 2] <- mean(c(SUMMARY$PC_pap))
df_glob_cor[2, 2] <- mean(c(SUMMARY$PC_pasp))
df_glob_cor[3, 2] <- mean(c(SUMMARY$PC_mapp))
df_glob_cor[4, 2] <- mean(c(SUMMARY$PC_mapsp))
df_glob_cor[5, 2] <- mean(c(SUMMARY$PC_egap))
df_glob_cor[6, 2] <- mean(c(SUMMARY$PC_egasp))
df_glob_cor[7, 2] <- mean(c(SUMMARY$PC_kpp))
df_glob_cor[8, 2] <- mean(c(SUMMARY$PC_kpsp))
df_glob_cor[9, 2] <- mean(c(SUMMARY$PC_kppm))
df_glob_cor[10, 2] <- mean(c(SUMMARY$PC_kpspm))
df_glob_cor[11, 2] <- mean(c(SUMMARY$PC_kpd))
df_glob_cor[1, 3] <- mean(c(SUMMARY$MBE_pap))
df_glob_cor[2, 3] <- mean(c(SUMMARY$MBE_pasp))
df_glob_cor[3, 3] <- mean(c(SUMMARY$MBE_mapp))
df_glob_cor[4, 3] <- mean(c(SUMMARY$MBE_mapsp))
df_glob_cor[5, 3] <- mean(c(SUMMARY$MBE_egap))
df_glob_cor[6, 3] <- mean(c(SUMMARY$MBE_egasp))
df_glob_cor[7, 3] <- mean(c(SUMMARY$MBE_kpp))
df_glob_cor[8, 3] <- mean(c(SUMMARY$MBE_kpsp))
df_glob_cor[9, 3] <- mean(c(SUMMARY$MBE_kppm))
df_glob_cor[10, 3] <- mean(c(SUMMARY$MBE_kpspm))
df_glob_cor[11, 3] <- mean(c(SUMMARY$MBE_kpd))
df_glob_cor[1, 4] <- mean(c(SUMMARY$MAE_pap))
df_glob_cor[2, 4] <- mean(c(SUMMARY$MAE_pasp))
df_glob_cor[3, 4] <- mean(c(SUMMARY$MAE_mapp))
df_glob_cor[4, 4] <- mean(c(SUMMARY$MAE_mapsp))
df_glob_cor[5, 4] <- mean(c(SUMMARY$MAE_egap))
df_glob_cor[6, 4] <- mean(c(SUMMARY$MAE_egasp))
df_glob_cor[7, 4] <- mean(c(SUMMARY$MAE_kpp))
df_glob_cor[8, 4] <- mean(c(SUMMARY$MAE_kpsp))
df_glob_cor[9, 4] <- mean(c(SUMMARY$MAE_kppm))
df_glob_cor[10, 4] <- mean(c(SUMMARY$MAE_kpspm))
df_glob_cor[11, 4] <- mean(c(SUMMARY$MAE_kpd))
kable(df_glob_cor, format = "html", table.attr = "class='table table-striped'") %>%
kable_styling(full_width = FALSE)| Metodo | mPC | mMBE | mMAE |
|---|---|---|---|
| PAp | 0.1895250 | -2.3398500 | 2.404825 |
| Pasp | 0.1895250 | -2.3405292 | 2.405379 |
| MAPp | 0.1765708 | 2.2614417 | 2.288592 |
| MAPsp | 0.1766792 | 2.2608625 | 2.288088 |
| EGAp | 0.4802333 | 0.5941125 | 1.214338 |
| EGAsp | 0.4799333 | 0.5956583 | 1.216133 |
| KP_mcp | 0.3177958 | -0.4288167 | 1.209742 |
| KP_mcsp | 0.3174750 | -0.4154083 | 1.206633 |
| KP_mcmp | 0.0527042 | -3.3984958 | 3.417421 |
| KP_mcmsp | 0.0527292 | -3.3977792 | 3.416729 |
| KP | 0.1317542 | -3.7780458 | 3.779104 |
Aparentemente no hay diferencias entre el tipo de matriz de correlación a usar (pendiente realizar prueba de hipótesis)
Resultados globales por factor
library(dplyr)
f1 <- SUMMARY %>% filter(Factor == "1")
f2 <- SUMMARY %>% filter(Factor == "2")
f4 <- SUMMARY %>% filter(Factor == "4")
f6 <- SUMMARY %>% filter(Factor == "6")
df_glob_fc <- data.frame(
Metodo = c("PA","MAP", "EGA","KP_mc","KP_mcm","KP"),
mPC1 = numeric(6),
mMBE1 = numeric(6),
mMAE1 = numeric(6),
mPC2 = numeric(6),
mMBE2 = numeric(6),
mMAE2 = numeric(6),
mPC4 = numeric(6),
mMBE4 = numeric(6),
mMAE4 = numeric(6),
mPC6 = numeric(6),
mMBE6 = numeric(6),
mMAE6 = numeric(6)
)
df_glob_fc[1, 2] <- mean(c(f1$PC_pasp, f1$PC_pasp))
df_glob_fc[2, 2] <- mean(c(f1$PC_mapp, f1$PC_mapsp))
df_glob_fc[3, 2] <- mean(c(f1$PC_egap, f1$PC_egasp))
df_glob_fc[4, 2] <- mean(c(f1$PC_kpp, f1$PC_kpsp))
df_glob_fc[5, 2] <- mean(c(f1$PC_kppm, f1$PC_kpspm))
df_glob_fc[6, 2] <- mean(c(f1$PC_kpd))
df_glob_fc[1, 3] <- mean(c(f1$MBE_pasp, f1$MBE_pasp))
df_glob_fc[2, 3] <- mean(c(f1$MBE_mapp, f1$MBE_mapsp))
df_glob_fc[3, 3] <- mean(c(f1$MBE_egap, f1$MBE_egasp))
df_glob_fc[4, 3] <- mean(c(f1$MBE_kpp, f1$MBE_kpsp))
df_glob_fc[5, 3] <- mean(c(f1$MBE_kppm, f1$MBE_kpspm))
df_glob_fc[6, 3] <- mean(c(f1$MBE_kpd))
df_glob_fc[1, 4] <- mean(c(f1$MAE_pasp, f1$MAE_pasp))
df_glob_fc[2, 4] <- mean(c(f1$MAE_mapp, f1$MAE_mapsp))
df_glob_fc[3, 4] <- mean(c(f1$MAE_egap, f1$MAE_egasp))
df_glob_fc[4, 4] <- mean(c(f1$MAE_kpp, f1$MAE_kpsp))
df_glob_fc[5, 4] <- mean(c(f1$MAE_kppm, f1$MAE_kpspm))
df_glob_fc[6, 4] <- mean(c(f1$MAE_kpd))
df_glob_fc[1, 5] <- mean(c(f2$PC_pasp, f2$PC_pasp))
df_glob_fc[2, 5] <- mean(c(f2$PC_mapp, f2$PC_mapsp))
df_glob_fc[3, 5] <- mean(c(f2$PC_egap, f2$PC_egasp))
df_glob_fc[4, 5] <- mean(c(f2$PC_kpp, f2$PC_kpsp))
df_glob_fc[5, 5] <- mean(c(f2$PC_kppm, f2$PC_kpspm))
df_glob_fc[6, 5] <- mean(c(f2$PC_kpd))
df_glob_fc[1, 6] <- mean(c(f2$MBE_pasp, f2$MBE_pasp))
df_glob_fc[2, 6] <- mean(c(f2$MBE_mapp, f2$MBE_mapsp))
df_glob_fc[3, 6] <- mean(c(f2$MBE_egap, f2$MBE_egasp))
df_glob_fc[4, 6] <- mean(c(f2$MBE_kpp, f2$MBE_kpsp))
df_glob_fc[5, 6] <- mean(c(f2$MBE_kppm, f2$MBE_kpspm))
df_glob_fc[6, 6] <- mean(c(f2$MBE_kpd))
df_glob_fc[1, 7] <- mean(c(f2$MAE_pasp, f2$MAE_pasp))
df_glob_fc[2, 7] <- mean(c(f2$MAE_mapp, f2$MAE_mapsp))
df_glob_fc[3, 7] <- mean(c(f2$MAE_egap, f2$MAE_egasp))
df_glob_fc[4, 7] <- mean(c(f2$MAE_kpp, f2$MAE_kpsp))
df_glob_fc[5, 7] <- mean(c(f2$MAE_kppm, f2$MAE_kpspm))
df_glob_fc[6, 7] <- mean(c(f2$MAE_kpd))
df_glob_fc[1, 8] <- mean(c(f4$PC_pasp, f4$PC_pasp))
df_glob_fc[2, 8] <- mean(c(f4$PC_mapp, f4$PC_mapsp))
df_glob_fc[3, 8] <- mean(c(f4$PC_egap, f4$PC_egasp))
df_glob_fc[4, 8] <- mean(c(f4$PC_kpp, f4$PC_kpsp))
df_glob_fc[5, 8] <- mean(c(f4$PC_kppm, f4$PC_kpspm))
df_glob_fc[6, 8] <- mean(c(f4$PC_kpd))
df_glob_fc[1, 9] <- mean(c(f4$MBE_pasp, f4$MBE_pasp))
df_glob_fc[2, 9] <- mean(c(f4$MBE_mapp, f4$MBE_mapsp))
df_glob_fc[3, 9] <- mean(c(f4$MBE_egap, f4$MBE_egasp))
df_glob_fc[4, 9] <- mean(c(f4$MBE_kpp, f4$MBE_kpsp))
df_glob_fc[5, 9] <- mean(c(f4$MBE_kppm, f4$MBE_kpspm))
df_glob_fc[6, 9] <- mean(c(f4$MBE_kpd))
df_glob_fc[1, 10] <- mean(c(f4$MAE_pasp, f4$MAE_pasp))
df_glob_fc[2, 10] <- mean(c(f4$MAE_mapp, f4$MAE_mapsp))
df_glob_fc[3, 10] <- mean(c(f4$MAE_egap, f4$MAE_egasp))
df_glob_fc[4, 10] <- mean(c(f4$MAE_kpp, f4$MAE_kpsp))
df_glob_fc[5, 10] <- mean(c(f4$MAE_kppm, f4$MAE_kpspm))
df_glob_fc[6, 10] <- mean(c(f4$MAE_kpd))
df_glob_fc[1, 11] <- mean(c(f6$PC_pasp, f6$PC_pasp))
df_glob_fc[2, 11] <- mean(c(f6$PC_mapp, f6$PC_mapsp))
df_glob_fc[3, 11] <- mean(c(f6$PC_egap, f6$PC_egasp))
df_glob_fc[4, 11] <- mean(c(f6$PC_kpp, f6$PC_kpsp))
df_glob_fc[5, 11] <- mean(c(f6$PC_kppm, f6$PC_kpspm))
df_glob_fc[6, 11] <- mean(c(f6$PC_kpd))
df_glob_fc[1, 12] <- mean(c(f6$MBE_pasp, f6$MBE_pasp))
df_glob_fc[2, 12] <- mean(c(f6$MBE_mapp, f6$MBE_mapsp))
df_glob_fc[3, 12] <- mean(c(f6$MBE_egap, f6$MBE_egasp))
df_glob_fc[4, 12] <- mean(c(f6$MBE_kpp, f6$MBE_kpsp))
df_glob_fc[5, 12] <- mean(c(f6$MBE_kppm, f6$MBE_kpspm))
df_glob_fc[6, 12] <- mean(c(f6$MBE_kpd))
df_glob_fc[1, 13] <- mean(c(f6$MAE_pasp, f6$MAE_pasp))
df_glob_fc[2, 13] <- mean(c(f6$MAE_mapp, f6$MAE_mapsp))
df_glob_fc[3, 13] <- mean(c(f6$MAE_egap, f6$MAE_egasp))
df_glob_fc[4, 13] <- mean(c(f6$MAE_kpp, f6$MAE_kpsp))
df_glob_fc[5, 13] <- mean(c(f6$MAE_kppm, f6$MAE_kpspm))
df_glob_fc[6, 13] <- mean(c(f6$MAE_kpd))
kable(df_glob_fc, format = "html", table.attr = "class='table table-striped'") %>%
kable_styling(full_width = FALSE)| Metodo | mPC1 | mMBE1 | mMAE1 | mPC2 | mMBE2 | mMAE2 | mPC4 | mMBE4 | mMAE4 | mPC6 | mMBE6 | mMAE6 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| PA | 0.2838750 | -1.1926667 | 1.2823333 | 0.2486667 | -1.6410694 | 1.6862917 | 0.1565000 | -2.5422083 | 2.582014 | 0.1319583 | -3.2209306 | 3.322181 |
| MAP | 0.4176250 | 0.5823750 | 0.5823750 | 0.2156042 | 1.2766042 | 1.2793264 | 0.1438681 | 2.4497639 | 2.477236 | 0.0900694 | 3.6166806 | 3.677111 |
| EGA | 0.7602500 | -0.1967083 | 0.2745417 | 0.5706528 | 0.0518194 | 0.6594861 | 0.4446806 | 0.5891181 | 1.264299 | 0.3315278 | 1.4075833 | 2.035486 |
| KP_mc | 0.1873333 | -0.8165208 | 0.8165208 | 0.7632639 | -0.1657083 | 0.2674028 | 0.1519028 | -0.1563889 | 1.340667 | 0.0811736 | -0.8127708 | 2.147049 |
| KP_mcm | 0.0069792 | -2.1570000 | 2.1570000 | 0.0881458 | -2.2589097 | 2.2869097 | 0.0475278 | -3.5900764 | 3.610118 | 0.0377222 | -4.7591389 | 4.774222 |
| KP | 0.3167500 | -0.7737083 | 0.7737083 | 0.2667639 | -1.6999444 | 1.7000833 | 0.0508889 | -4.1320556 | 4.133278 | 0.0159444 | -6.5035833 | 6.505750 |
Visualmente el número de factores si parece incidir.
Se cruza ahora la anterior información con el tipo de matriz de correlación
library(dplyr)
f1m1 <- SUMMARY %>% filter(Factor == "1") %>% filter(Matrix =="1")
f1m2 <- SUMMARY %>% filter(Factor == "1") %>% filter(Matrix =="2")
f2m1 <- SUMMARY %>% filter(Factor == "2") %>% filter(Matrix =="1")
f2m2 <- SUMMARY %>% filter(Factor == "2") %>% filter(Matrix =="2")
f4m1 <- SUMMARY %>% filter(Factor == "4") %>% filter(Matrix =="1")
f4m2 <- SUMMARY %>% filter(Factor == "4") %>% filter(Matrix =="2")
f6m1 <- SUMMARY %>% filter(Factor == "6") %>% filter(Matrix =="1")
f6m2 <- SUMMARY %>% filter(Factor == "6") %>% filter(Matrix =="2")
df_glob_fc_m <- data.frame(
Metodo = c("PA","PA" ,"MAP","MAP","EGA","EGA","KP_mc", "KP_mc",
"KP_mcm","KP_mcm","KP", "KP"),
matrix = c("1", "2", "1", "2", "1", "2", "1", "2", "1", "2", "1", "2"),
mPC1 = numeric(12),
mMBE1 = numeric(12),
mMAE1 = numeric(12),
mPC2 = numeric(12),
mMBE2 = numeric(12),
mMAE2 = numeric(12),
mPC4 = numeric(12),
mMBE4 = numeric(12),
mMAE4 = numeric(12),
mPC6 = numeric(12),
mMBE6 = numeric(12),
mMAE6 = numeric(12)
)
df_glob_fc_m[1, 3] <- mean(c(f1m1$PC_pasp, f1m1$PC_pasp))
df_glob_fc_m[2, 3] <- mean(c(f1m2$PC_pasp, f1m2$PC_pasp))
df_glob_fc_m[3, 3] <- mean(c(f1m1$PC_mapp, f1m1$PC_mapsp))
df_glob_fc_m[4, 3] <- mean(c(f1m2$PC_mapp, f1m2$PC_mapsp))
df_glob_fc_m[5, 3] <- mean(c(f1m1$PC_egap, f1m1$PC_egasp))
df_glob_fc_m[6, 3] <- mean(c(f1m2$PC_egap, f1m2$PC_egasp))
df_glob_fc_m[7, 3] <- mean(c(f1m1$PC_kpp, f1m1$PC_kpsp))
df_glob_fc_m[8, 3] <- mean(c(f1m2$PC_kpp, f1m2$PC_kpsp))
df_glob_fc_m[9, 3] <- mean(c(f1m1$PC_kppm, f1m1$PC_kpspm))
df_glob_fc_m[10, 3] <- mean(c(f1m2$PC_kppm, f1m2$PC_kpspm))
df_glob_fc_m[11, 3] <- mean(c(f1m1$PC_kpd))
df_glob_fc_m[12, 3] <- mean(c(f1m2$PC_kpd))
df_glob_fc_m[1, 4] <- mean(c(f1m1$MBE_pasp, f1m1$MBE_pasp))
df_glob_fc_m[2, 4] <- mean(c(f1m2$MBE_pasp, f1m2$MBE_pasp))
df_glob_fc_m[3, 4] <- mean(c(f1m1$MBE_mapp, f1m1$MBE_mapsp))
df_glob_fc_m[4, 4] <- mean(c(f1m2$MBE_mapp, f1m2$MBE_mapsp))
df_glob_fc_m[5, 4] <- mean(c(f1m1$MBE_egap, f1m1$MBE_egasp))
df_glob_fc_m[6, 4] <- mean(c(f1m2$MBE_egap, f1m2$MBE_egasp))
df_glob_fc_m[7, 4] <- mean(c(f1m1$MBE_kpp, f1m1$MBE_kpsp))
df_glob_fc_m[8, 4] <- mean(c(f1m2$MBE_kpp, f1m2$MBE_kpsp))
df_glob_fc_m[9, 4] <- mean(c(f1m1$MBE_kppm, f1m1$MBE_kpspm))
df_glob_fc_m[10, 4] <- mean(c(f1m2$MBE_kppm, f1m2$MBE_kpspm))
df_glob_fc_m[11, 4] <- mean(c(f1m1$MBE_kpd))
df_glob_fc_m[12, 4] <- mean(c(f1m2$MBE_kpd))
df_glob_fc_m[1, 5] <- mean(c(f1m1$MAE_pasp, f1m1$MAE_pasp))
df_glob_fc_m[2, 5] <- mean(c(f1m2$MAE_pasp, f1m2$MAE_pasp))
df_glob_fc_m[3, 5] <- mean(c(f1m1$MAE_mapp, f1m1$MAE_mapsp))
df_glob_fc_m[4, 5] <- mean(c(f1m2$MAE_mapp, f1m2$MAE_mapsp))
df_glob_fc_m[5, 5] <- mean(c(f1m1$MAE_egap, f1m1$MAE_egasp))
df_glob_fc_m[6, 5] <- mean(c(f1m2$MAE_egap, f1m2$MAE_egasp))
df_glob_fc_m[7, 5] <- mean(c(f1m1$MAE_kpp, f1m1$MAE_kpsp))
df_glob_fc_m[8, 5] <- mean(c(f1m2$MAE_kpp, f1m2$MAE_kpsp))
df_glob_fc_m[9, 5] <- mean(c(f1m1$MAE_kppm, f1m1$MAE_kpspm))
df_glob_fc_m[10,5] <- mean(c(f1m2$MAE_kppm, f1m2$MAE_kpspm))
df_glob_fc_m[11, 5] <- mean(c(f1m1$MAE_kpd))
df_glob_fc_m[12, 5] <- mean(c(f1m2$MAE_kpd))
df_glob_fc_m[1, 6] <- mean(c(f2m1$PC_pasp, f2m1$PC_pasp))
df_glob_fc_m[2, 6] <- mean(c(f2m2$PC_pasp, f2m2$PC_pasp))
df_glob_fc_m[3, 6] <- mean(c(f2m1$PC_mapp, f2m1$PC_mapsp))
df_glob_fc_m[4, 6] <- mean(c(f2m2$PC_mapp, f2m2$PC_mapsp))
df_glob_fc_m[5, 6] <- mean(c(f2m1$PC_egap, f2m1$PC_egasp))
df_glob_fc_m[6, 6] <- mean(c(f2m2$PC_egap, f2m2$PC_egasp))
df_glob_fc_m[7, 6] <- mean(c(f2m1$PC_kpp, f2m1$PC_kpsp))
df_glob_fc_m[8, 6] <- mean(c(f2m2$PC_kpp, f2m2$PC_kpsp))
df_glob_fc_m[9, 6] <- mean(c(f2m1$PC_kppm, f2m1$PC_kpspm))
df_glob_fc_m[10,6] <- mean(c(f2m2$PC_kppm, f2m2$PC_kpspm))
df_glob_fc_m[11, 6] <- mean(c(f2m1$PC_kpd))
df_glob_fc_m[12, 6] <- mean(c(f2m2$PC_kpd))
df_glob_fc_m[1, 7] <- mean(c(f2m1$MBE_pasp, f2m1$MBE_pasp))
df_glob_fc_m[2, 7] <- mean(c(f2m2$MBE_pasp, f2m2$MBE_pasp))
df_glob_fc_m[3, 7] <- mean(c(f2m1$MBE_mapp, f2m1$MBE_mapsp))
df_glob_fc_m[4, 7] <- mean(c(f2m2$MBE_mapp, f2m2$MBE_mapsp))
df_glob_fc_m[5, 7] <- mean(c(f2m1$MBE_egap, f2m1$MBE_egasp))
df_glob_fc_m[6, 7] <- mean(c(f2m2$MBE_egap, f2m2$MBE_egasp))
df_glob_fc_m[7, 7] <- mean(c(f2m1$MBE_kpp, f2m1$MBE_kpsp))
df_glob_fc_m[8, 7] <- mean(c(f2m2$MBE_kpp, f2m2$MBE_kpsp))
df_glob_fc_m[9, 7] <- mean(c(f2m1$MBE_kppm, f2m1$MBE_kpspm))
df_glob_fc_m[10, 7] <- mean(c(f2m2$MBE_kppm, f2m2$MBE_kpspm))
df_glob_fc_m[11, 7] <- mean(c(f2m1$MBE_kpd))
df_glob_fc_m[12, 7] <- mean(c(f2m2$MBE_kpd))
df_glob_fc_m[1, 8] <- mean(c(f2m1$MAE_pasp, f2m1$MAE_pasp))
df_glob_fc_m[2, 8] <- mean(c(f2m2$MAE_pasp, f2m2$MAE_pasp))
df_glob_fc_m[3, 8] <- mean(c(f2m1$MAE_mapp, f2m1$MAE_mapsp))
df_glob_fc_m[4, 8] <- mean(c(f2m2$MAE_mapp, f2m2$MAE_mapsp))
df_glob_fc_m[5, 8] <- mean(c(f2m1$MAE_egap, f2m1$MAE_egasp))
df_glob_fc_m[6, 8] <- mean(c(f2m2$MAE_egap, f2m2$MAE_egasp))
df_glob_fc_m[7, 8] <- mean(c(f2m1$MAE_kpp, f2m1$MAE_kpsp))
df_glob_fc_m[8, 8] <- mean(c(f2m2$MAE_kpp, f2m2$MAE_kpsp))
df_glob_fc_m[9, 8] <- mean(c(f2m1$MAE_kppm, f2m1$MAE_kpspm))
df_glob_fc_m[10, 8] <- mean(c(f2m2$MAE_kppm, f2m2$MAE_kpspm))
df_glob_fc_m[11, 8] <- mean(c(f2m1$MAE_kpd))
df_glob_fc_m[12, 8] <- mean(c(f2m2$MAE_kpd))
df_glob_fc_m[1, 9] <- mean(c(f4m1$PC_pasp, f4m1$PC_pasp))
df_glob_fc_m[2, 9] <- mean(c(f4m2$PC_pasp, f4m2$PC_pasp))
df_glob_fc_m[3, 9] <- mean(c(f4m1$PC_mapp, f4m1$PC_mapsp))
df_glob_fc_m[4, 9] <- mean(c(f4m2$PC_mapp, f4m2$PC_mapsp))
df_glob_fc_m[5, 9] <- mean(c(f4m1$PC_egap, f4m1$PC_egasp))
df_glob_fc_m[6, 9] <- mean(c(f4m2$PC_egap, f4m2$PC_egasp))
df_glob_fc_m[7, 9] <- mean(c(f4m1$PC_kpp, f4m1$PC_kpsp))
df_glob_fc_m[8, 9] <- mean(c(f4m2$PC_kpp, f4m2$PC_kpsp))
df_glob_fc_m[9, 9] <- mean(c(f4m1$PC_kppm, f4m1$PC_kpspm))
df_glob_fc_m[10, 9] <- mean(c(f4m2$PC_kppm, f4m2$PC_kpspm))
df_glob_fc_m[11, 9] <- mean(c(f4m1$PC_kpd))
df_glob_fc_m[12, 9] <- mean(c(f4m2$PC_kpd))
df_glob_fc_m[1, 10] <- mean(c(f4m1$MBE_pasp, f4m1$MBE_pasp))
df_glob_fc_m[2, 10] <- mean(c(f4m2$MBE_pasp, f4m2$MBE_pasp))
df_glob_fc_m[3, 10] <- mean(c(f4m1$MBE_mapp, f4m1$MBE_mapsp))
df_glob_fc_m[4, 10] <- mean(c(f4m2$MBE_mapp, f4m2$MBE_mapsp))
df_glob_fc_m[5, 10] <- mean(c(f4m1$MBE_egap, f4m1$MBE_egasp))
df_glob_fc_m[6, 10] <- mean(c(f4m2$MBE_egap, f4m2$MBE_egasp))
df_glob_fc_m[7, 10] <- mean(c(f4m1$MBE_kpp, f4m1$MBE_kpsp))
df_glob_fc_m[8, 10] <- mean(c(f4m2$MBE_kpp, f4m2$MBE_kpsp))
df_glob_fc_m[9, 10] <- mean(c(f4m1$MBE_kppm, f4m1$MBE_kpspm))
df_glob_fc_m[10, 10] <- mean(c(f4m2$MBE_kppm, f4m2$MBE_kpspm))
df_glob_fc_m[11, 10] <- mean(c(f4m1$MBE_kpd))
df_glob_fc_m[12, 10] <- mean(c(f4m2$MBE_kpd))
df_glob_fc_m[1, 11] <- mean(c(f4m1$MAE_pasp, f4m1$MAE_pasp))
df_glob_fc_m[2, 11] <- mean(c(f4m2$MAE_pasp, f4m2$MAE_pasp))
df_glob_fc_m[3, 11] <- mean(c(f4m1$MAE_mapp, f4m1$MAE_mapsp))
df_glob_fc_m[4, 11] <- mean(c(f4m2$MAE_mapp, f4m2$MAE_mapsp))
df_glob_fc_m[5, 11] <- mean(c(f4m1$MAE_egap, f4m1$MAE_egasp))
df_glob_fc_m[6, 11] <- mean(c(f4m2$MAE_egap, f4m2$MAE_egasp))
df_glob_fc_m[7, 11] <- mean(c(f4m1$MAE_kpp, f4m1$MAE_kpsp))
df_glob_fc_m[8, 11] <- mean(c(f4m2$MAE_kpp, f4m2$MAE_kpsp))
df_glob_fc_m[9, 11] <- mean(c(f4m1$MAE_kppm, f4m1$MAE_kpspm))
df_glob_fc_m[10, 11] <- mean(c(f4m2$MAE_kppm, f4m2$MAE_kpspm))
df_glob_fc_m[11, 11] <- mean(c(f4m1$MAE_kpd))
df_glob_fc_m[12, 11] <- mean(c(f4m2$MAE_kpd))
df_glob_fc_m[1, 12] <- mean(c(f6m1$PC_pasp, f6m1$PC_pasp))
df_glob_fc_m[2, 12] <- mean(c(f6m2$PC_pasp, f6m2$PC_pasp))
df_glob_fc_m[3, 12] <- mean(c(f6m1$PC_mapp, f6m1$PC_mapsp))
df_glob_fc_m[4, 12] <- mean(c(f6m2$PC_mapp, f6m2$PC_mapsp))
df_glob_fc_m[5, 12] <- mean(c(f6m1$PC_egap, f6m1$PC_egasp))
df_glob_fc_m[6, 12] <- mean(c(f6m2$PC_egap, f6m2$PC_egasp))
df_glob_fc_m[7, 12] <- mean(c(f6m1$PC_kpp, f6m1$PC_kpsp))
df_glob_fc_m[8, 12] <- mean(c(f6m2$PC_kpp, f6m2$PC_kpsp))
df_glob_fc_m[9, 12] <- mean(c(f6m1$PC_kppm, f6m1$PC_kpspm))
df_glob_fc_m[10, 12] <- mean(c(f6m2$PC_kppm, f6m2$PC_kpspm))
df_glob_fc_m[11, 12] <- mean(c(f6m1$PC_kpd))
df_glob_fc_m[12, 12] <- mean(c(f6m2$PC_kpd))
df_glob_fc_m[1, 13] <- mean(c(f6m1$MBE_pasp, f6m1$MBE_pasp))
df_glob_fc_m[2, 13] <- mean(c(f6m2$MBE_pasp, f6m2$MBE_pasp))
df_glob_fc_m[3, 13] <- mean(c(f6m1$MBE_mapp, f6m1$MBE_mapsp))
df_glob_fc_m[4, 13] <- mean(c(f6m2$MBE_mapp, f6m2$MBE_mapsp))
df_glob_fc_m[5, 13] <- mean(c(f6m1$MBE_egap, f6m1$MBE_egasp))
df_glob_fc_m[6, 13] <- mean(c(f6m2$MBE_egap, f6m2$MBE_egasp))
df_glob_fc_m[7, 13] <- mean(c(f6m1$MBE_kpp, f6m1$MBE_kpsp))
df_glob_fc_m[8, 13] <- mean(c(f6m2$MBE_kpp, f6m2$MBE_kpsp))
df_glob_fc_m[9, 13] <- mean(c(f6m1$MBE_kppm, f6m1$MBE_kpspm))
df_glob_fc_m[10, 13] <- mean(c(f6m2$MBE_kppm, f6m2$MBE_kpspm))
df_glob_fc_m[11, 13] <- mean(c(f6m1$MBE_kpd))
df_glob_fc_m[12, 13] <- mean(c(f6m2$MBE_kpd))
df_glob_fc_m[1, 14] <- mean(c(f6m1$MAE_pasp, f6m1$MAE_pasp))
df_glob_fc_m[2, 14] <- mean(c(f6m2$MAE_pasp, f6m2$MAE_pasp))
df_glob_fc_m[3, 14] <- mean(c(f6m1$MAE_mapp, f6m1$MAE_mapsp))
df_glob_fc_m[4, 14] <- mean(c(f6m2$MAE_mapp, f6m2$MAE_mapsp))
df_glob_fc_m[5, 14] <- mean(c(f6m1$MAE_egap, f6m1$MAE_egasp))
df_glob_fc_m[6, 14] <- mean(c(f6m2$MAE_egap, f6m2$MAE_egasp))
df_glob_fc_m[7, 14] <- mean(c(f6m1$MAE_kpp, f6m1$MAE_kpsp))
df_glob_fc_m[8, 14] <- mean(c(f6m2$MAE_kpp, f6m2$MAE_kpsp))
df_glob_fc_m[9, 14] <- mean(c(f6m1$MAE_kppm, f6m1$MAE_kpspm))
df_glob_fc_m[10, 14] <- mean(c(f6m2$MAE_kppm, f6m2$MAE_kpspm))
df_glob_fc_m[11, 14] <- mean(c(f6m1$MAE_kpd))
df_glob_fc_m[12, 14] <- mean(c(f6m2$MAE_kpd))
kable(df_glob_fc_m, format = "html", table.attr = "class='table table-striped'") %>%
kable_styling(full_width = FALSE)| Metodo | matrix | mPC1 | mMBE1 | mMAE1 | mPC2 | mMBE2 | mMAE2 | mPC4 | mMBE4 | mMAE4 | mPC6 | mMBE6 | mMAE6 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| PA | 1 | 0.2518333 | -1.3230000 | 1.4208333 | 0.2558056 | -1.6256944 | 1.6747500 | 0.1499167 | -2.7279444 | 2.769111 | 0.1150000 | -3.3459167 | 3.450972 |
| PA | 2 | 0.3159167 | -1.0623333 | 1.1438333 | 0.2415278 | -1.6564444 | 1.6978333 | 0.1630833 | -2.3564722 | 2.394917 | 0.1489167 | -3.0959444 | 3.193389 |
| MAP | 1 | 0.3305417 | 0.6694583 | 0.6694583 | 0.2157500 | 1.2745556 | 1.2769444 | 0.1334722 | 2.5543194 | 2.579708 | 0.0873889 | 3.6757778 | 3.734778 |
| MAP | 2 | 0.5047083 | 0.4952917 | 0.4952917 | 0.2154583 | 1.2786528 | 1.2817083 | 0.1542639 | 2.3452083 | 2.374764 | 0.0927500 | 3.5575833 | 3.619444 |
| EGA | 1 | 0.6947917 | -0.2213333 | 0.3496667 | 0.5721389 | 0.0486111 | 0.6574722 | 0.4225694 | 0.7485972 | 1.377597 | 0.2980833 | 1.6560417 | 2.284708 |
| EGA | 2 | 0.8257083 | -0.1720833 | 0.1994167 | 0.5691667 | 0.0550278 | 0.6615000 | 0.4667917 | 0.4296389 | 1.151000 | 0.3649722 | 1.1591250 | 1.786264 |
| KP_mc | 1 | 0.1789583 | -0.8260000 | 0.8260000 | 0.7641250 | -0.1632222 | 0.2675278 | 0.1414167 | -0.2540833 | 1.426917 | 0.0956806 | -0.8735139 | 2.113903 |
| KP_mc | 2 | 0.1957083 | -0.8070417 | 0.8070417 | 0.7624028 | -0.1681944 | 0.2672778 | 0.1623889 | -0.0586944 | 1.254417 | 0.0666667 | -0.7520278 | 2.180194 |
| KP_mcm | 1 | 0.0062500 | -2.2061667 | 2.2061667 | 0.0905972 | -2.2542222 | 2.2851111 | 0.0453889 | -3.7278611 | 3.747306 | 0.0308194 | -4.8722639 | 4.879236 |
| KP_mcm | 2 | 0.0077083 | -2.1078333 | 2.1078333 | 0.0856944 | -2.2635972 | 2.2887083 | 0.0496667 | -3.4522917 | 3.472931 | 0.0446250 | -4.6460139 | 4.669208 |
| KP | 1 | 0.2927500 | -0.8478333 | 0.8478333 | 0.2712222 | -1.6914722 | 1.6917500 | 0.0522778 | -4.2563611 | 4.257694 | 0.0101944 | -6.6181944 | 6.619528 |
| KP | 2 | 0.3407500 | -0.6995833 | 0.6995833 | 0.2623056 | -1.7084167 | 1.7084167 | 0.0495000 | -4.0077500 | 4.008861 | 0.0216944 | -6.3889722 | 6.391972 |
Especialmente en el caso undimensional la matriz de correlación parece afectar la extracción de factores, auqnue también en otros métodos para afectar la detección de factores (6 factores)
Intento de grafico de la desviacion del MBE para el análisis paralelo con el MAE como error
library(ggplot2)
df_glob_fc_m$matrix <- as.factor(df_glob_fc_m$matrix)
df_glob_fc_m$Metodo <- as.factor(df_glob_fc_m$Metodo )
ggplot(df_glob_fc_m, aes(x = matrix, y = mMBE1, color = Metodo)) +
geom_point(position = position_dodge(width = 0.5)) +
geom_errorbar(aes(ymin = mMBE1 - mMAE1, ymax = mMBE1 + mMAE1),
width = 0.2, position = position_dodge(width = 0.5)) +
labs(title = "Desviación de MBE con MAE como Error factor 1",
x = "Tipo de matriz",
y = "MBE (con MAE como barras de error)") +
theme_minimal()
# 2 factores
ggplot(df_glob_fc_m, aes(x = matrix, y = mMBE2, color = Metodo)) +
geom_point(position = position_dodge(width = 0.5)) +
geom_errorbar(aes(ymin = mMBE2 - mMAE2, ymax = mMBE2 + mMAE2),
width = 0.2, position = position_dodge(width = 0.5)) +
labs(title = "Desviación de MBE con MAE como Error factor 2",
x = "Tipo de matriz",
y = "MBE (con MAE como barras de error)") +
theme_minimal()
# 4 factores
ggplot(df_glob_fc_m, aes(x = matrix, y = mMBE4, color = Metodo)) +
geom_point(position = position_dodge(width = 0.5)) +
geom_errorbar(aes(ymin = mMBE4 - mMAE4, ymax = mMBE4 + mMAE4),
width = 0.2, position = position_dodge(width = 0.5)) +
labs(title = "Desviación de MBE con MAE como Error factor 4",
x = "Tipo de matriz",
y = "MBE (con MAE como barras de error)") +
theme_minimal()
6 factores
ggplot(df_glob_fc_m, aes(x = matrix, y = mMBE6, color = Metodo)) +
geom_point(position = position_dodge(width = 0.5)) +
geom_errorbar(aes(ymin = mMBE6 - mMAE6, ymax = mMBE6 + mMAE6),
width = 0.2, position = position_dodge(width = 0.5)) +
labs(title = "Desviación de MBE con MAE como Error factor 6",
x = "Tipo de matriz",
y = "MBE (con MAE como barras de error)") +
theme_minimal()