R/Old_code/ggplot_20_variant_results.R
In langfob/bdpg: Package For Building and Testing Biodiversity Problem Generators
Defines functions
dummy
#===============================================================================
#
# ggplot_20_variant_results.R
#
# This function is not meant to be called.
# It's only here as a dummy wrapper to keep R from executing this code when
# it builds the package.
# The code here is currently just a bunch of hacking around to do learn
# something about plotting with ggplot and at the same time, look at some
# very preliminary bdpg outputs from getting things working on nectar.
# It will probably evolve into real plotting code eventually, but right now,
# it's just a graveyard for code as I explore.
#
#===============================================================================
load_input_data <- function (rs_name, base_path, suffix,
num_sets = 4,
num_variants_per_set = 20,
num_combined_err_labels = 10)
{
infile = paste0 (base_path, rs_name, suffix)
msa_dt = read.csv (infile, header=TRUE, stringsAsFactors = FALSE)
msa_tib = as.tibble (msa_dt)
sorted_msa_tib = arrange (msa_tib, rsp_combined_err_label)
# Now that they are sorted by the 10 combined error labels,
# give an index to each rsrun within each of the combined error labels
# so that they can be spread out by index across a single facet plot.
# The index has no meaning other than to give it a unique identifier
# to spread along the x axis within each facet.
num_indices_per_combined_err_label =
num_sets * num_variants_per_set * num_combined_err_labels
sorted_msa_tib$idx = rep (1:num_indices_per_combined_err_label,
num_combined_err_labels)
unique_out_errs = unique (sorted_msa_tib$rsr_COR_euc_out_err_frac)
length (unique_out_errs)
sorted_unique_out_errs = sort (unique_out_errs)
cat ("\n\nhead (sorted_unique_OUT_errs) = \n")
print (head (sorted_unique_out_errs))
cat ("\ntail (sorted_unique_OUT_errs) = \n")
print (tail (sorted_unique_out_errs))
cat ("\n-----------------------")
unique_in_errs = unique (sorted_msa_tib$rsp_euc_realized_Ftot_and_cost_in_err_frac)
length (unique_in_errs)
sorted_unique_in_errs = sort (unique_in_errs)
cat ("\n\nhead (sorted_unique_IN_errs) = \n")
print (head (sorted_unique_in_errs))
cat ("\ntail (sorted_unique_IN_errs) = \n")
print (tail (sorted_unique_in_errs))
cat ("\n-----------------------")
return (sorted_msa_tib)
}
#===============================================================================
#' @importFrom stats mad quantile sd
compute_and_print_na_inf_stats <- function (rs_name,
out_err,
inp_err,
err_mag,
print_heading_str)
{
cat ("\n\n------------------------------------------------")
cat (paste0 ("\n\n----- ", print_heading_str, " -----"))
#----------
num_inp_na = length (which (is.na (inp_err)))
cat ("\n\nnum_inp_na for err_mag = ", num_inp_na)
num_inp_inf = length (which (is.infinite (inp_err)))
cat ("\nnum_inp_inf for err_mag = ", num_inp_inf)
num_inp_0 = length (which (inp_err == 0))
cat ("\nnum_inp_0 for err_mag = ", num_inp_0, "\n\n")
#----------
num_out_na = length (which (is.na (out_err)))
cat ("\n\nnum_out_na for err_mag = ", num_out_na)
num_out_inf = length (which (is.infinite (out_err)))
cat ("\nnum_out_inf for err_mag = ", num_out_inf)
num_out_0 = length (which (inp_err == 0))
cat ("\nnum_out_0 for err_mag = ", num_out_0, "\n\n")
#----------
num_mag_na = length (which (is.na (err_mag)))
cat ("\n\nnum_mag_na for err_mag = ", num_mag_na)
num_mag_inf = length (which (is.infinite (err_mag)))
cat ("\nnum_mag_inf for err_mag = ", num_mag_inf)
num_mag_0 = length (which (inp_err == 0))
cat ("\nnum_mag_0 for err_mag = ", num_mag_0, "\n\n")
#----------
# Set infinite values to NA.
err_mag [is.infinite (err_mag)] = NA
mean_err_mag = mean (err_mag, na.rm = TRUE)
sd_err_mag = sd (err_mag, na.rm = TRUE)
median_err_mag = median (err_mag, na.rm = TRUE)
mad_err_mag = mad (err_mag, na.rm = TRUE)
cat (rs_name, " - Error magnification statistics (i.e., outErr/inErr):",
"\n\n mean = ", mean_err_mag,
"\n sd = ", sd_err_mag,
"\n\n median = ", median_err_mag,
"\n mad = ", mad_err_mag,
"\n\n", sep='')
cat ("Quantiles:\n")
print (quantile (err_mag, na.rm = TRUE))
#browser()
}
compute_and_print_mag_stats <- function (rs_name,
out_err,
inp_err
, rsr_UUID
, rsp_combined_err_label
)
{
err_mag = out_err / inp_err
inf_locs = which (is.infinite (err_mag))
if (length (inf_locs) > 0)
{
for (iii in 1:length (inf_locs))
{
cat ("\nInf err_mag [", inf_locs[iii],
"], rsr_UUID = ", rsr_UUID [inf_locs[iii]], ", ", rsp_combined_err_label [inf_locs[iii]], ": inp_err = ", inp_err[inf_locs[iii]], ",
out_err = ", out_err[inf_locs[iii]])
}
}
compute_and_print_na_inf_stats (rs_name,
out_err,
inp_err,
err_mag,
"FOR ALL DATA")
# cat ("\n\n------------------------------------------------")
# cat ("\n\n----- FOR ALL DATA -----")
# num_mag_na = length (which (is.na (err_mag)))
# cat ("\n\nnum_mag_na for err_mag = ", num_mag_na)
# num_inf = length (which (is.infinite (err_mag)))
# cat ("\nnum_inf for err_mag = ", num_inf, "\n\n")
# err_mag [is.infinite (err_mag)] = NA
#
# mean_err_mag = mean (err_mag, na.rm = TRUE)
# sd_err_mag = sd (err_mag, na.rm = TRUE)
#
# median_err_mag = median (err_mag, na.rm = TRUE)
# mad_err_mag = mad (err_mag, na.rm = TRUE)
#
# cat (rs_name, " - Error magnification statistics (i.e., outErr/inErr):",
# "\n\n mean = ", mean_err_mag,
# "\n sd = ", sd_err_mag,
# "\n\n median = ", median_err_mag,
# "\n mad = ", mad_err_mag,
# "\n\n", sep='')
#
# cat ("Quantiles:\n")
# print (quantile (err_mag, na.rm = TRUE))
#--------------------
# cat ("\n\n------------------------------------------------")
# cat ("\n\n----- FOR input error >= 0.01 -----")
ge1_indices = which (inp_err >= 0.01)
num_indices_ge1 = length (ge1_indices)
cat ("\n\nnum_indices_ge1 = ", num_indices_ge1)
err_mag = err_mag [ge1_indices]
compute_and_print_na_inf_stats (rs_name,
out_err,
inp_err,
err_mag,
"FOR input error >= 0.01")
# num_mag_na = length (which (is.na (err_mag)))
# cat ("\n\nnum_mag_na for err_mag = ", num_mag_na)
#
# num_inf = length (which (is.infinite (err_mag)))
# cat ("\nnum_inf for err_mag = ", num_inf, "\n\n")
# err_mag [is.infinite (err_mag)] = NA
#
# mean_err_mag = mean (err_mag, na.rm = TRUE)
# sd_err_mag = sd (err_mag, na.rm = TRUE)
#
# median_err_mag = median (err_mag, na.rm = TRUE)
# mad_err_mag = mad (err_mag, na.rm = TRUE)
#
# cat (rs_name, " - Error magnification statistics (i.e., outErr/inErr):",
# "\n\n mean = ", mean_err_mag,
# "\n sd = ", sd_err_mag,
# "\n\n median = ", median_err_mag,
# "\n mad = ", mad_err_mag,
# "\n\n", sep='')
#
# cat ("Quantiles:\n")
# print (quantile (err_mag, na.rm = TRUE))
}
#===============================================================================
gg_multiple_err_amts <- function (rs_name, base_path, suffix,
num_sets = 4,
num_variants_per_set = 20,
num_combined_err_labels = 10)
{
sorted_msa_tib = load_input_data (rs_name, base_path, suffix)
compute_and_print_mag_stats (rs_name,
sorted_msa_tib$rsr_COR_euc_out_err_frac,
sorted_msa_tib$rsp_euc_realized_Ftot_and_cost_in_err_frac
, sorted_msa_tib$rsr_UUID
, sorted_msa_tib$rsp_combined_err_label)
return (sorted_msa_tib)
}
#===============================================================================
gg_eucInTot_vs_eucOutTot_all_on_1 <- function (rs_name,
sorted_msa_tib,
ref_y = 0)
{
ggplot (data = sorted_msa_tib) +
aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac,
y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str) +
geom_point() +
scale_color_manual(breaks = c("Base", "Wrap"), values=c("red", "blue")) +
stat_smooth(method = "lm", col = "green") +
# ggplot (data = sorted_msa_tib) +
# geom_point (mapping = ) +
# #scale_color_viridis(discrete=TRUE) +
# scale_color_manual(breaks = c("Base", "Wrap"), values=c("red", "blue")) +
#
ggtitle (paste0 (rs_name, " - Total input error vs. Total output error")) +
theme(plot.title = element_text(hjust = 0.5)) + # To center the title
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
}
#===============================================================================
gg_eucInTot_vs_eucOutTot_facetted_by_err_label <- function (rs_name,
sorted_msa_tib,
ref_y = 0)
{
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac,
y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
#scale_color_viridis(discrete=TRUE) +
#scale_color_brewer(palette="Set1") +
scale_color_manual(breaks = c("Base", "Wrap"), values=c("red", "blue")) +
ggtitle (paste0 (rs_name, " - Total input error vs. Total output error by Error Class")) +
theme(plot.title = element_text(hjust = 0.5)) + # To center the title
facet_wrap (~ rsp_combined_err_label, nrow = 5) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
}
#===============================================================================
dummy <- function()
{
#library (tidyverse)
base_path = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/"
suffix = ".combined_results.csv"
rs_name = "Gurobi"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "Marxan_SA"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "ZL_Backward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "ZL_Forward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "UR_Backward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "UR_Forward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "SR_Backward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "SR_Forward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
}
#===============================================================================
#===============================================================================
#===============================================================================
gg_20 <- function ()
{
#===============================================================================
#===============================================================================
#===============================================================================
# 4 sets of 20 outputs from 100 gen_20_basic_variants() runs
# on nectar and copied to glass.
#===============================================================================
### Load necessary libraries.
#library (tidyverse)
# rs_name = "Gurobi"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/Gurobi.combined_results.csv"
infile = paste0 (base_path, rs_name, suffix)
# rs_name = "Marxan_SA"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/Marxan_SA.combined_results.csv"
# rs_name = "ZL_Backward"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/ZL_Backward.combined_results.csv"
# rs_name = "ZL_Forward"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/ZL_Forward.combined_results.csv"
# rs_name = "SR_Forward"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/SR_Forward.combined_results.csv"
# rs_name = "UR_Forward"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/UR_Forward.combined_results.csv"
# rs_name = "UR_Backward"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/UR_Backward.combined_results.csv"
rs_name = "SR_Backward"
infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/SR_Backward.combined_results.csv"
ref_y = 0
msa_dt = read.csv (infile, header=TRUE, stringsAsFactors = FALSE)
msa_tib = as.tibble (msa_dt)
sorted_msa_tib = arrange (msa_tib, rsp_combined_err_label)
# Now that they are sorted by the 10 combined error labels,
# give an index to each rsrun within each of the combined error labels
# so that they can be spread out by index across a single facet plot.
# The index has no meaning other than to give it a unique identifier
# to spread along the x axis within each facet.
sorted_msa_tib$idx = rep (1:800, 10) #1:dim(sorted_msa_tib)[1] # For the 4 combined sets 0f 100 runs each
#----------
# NOTE: the result of selecting for NA gives 201 instead of 200 results.
# I would expect 200 since there are 200 correct values that have no
# input error. Not sure what 1 extra NA is from.
err_mag_euc_Ftot_in_vs_euc_Ftot_out =
sorted_msa_tib$rsr_COR_euc_out_err_frac /
sorted_msa_tib$rsp_euc_realized_Ftot_and_cost_in_err_frac
err_mag_euc_Ftot_in_vs_euc_Ftot_out [is.na (err_mag_euc_Ftot_in_vs_euc_Ftot_out)] = 0
mean_err_mag_euc_Ftot_in_vs_euc_Ftot_out = mean (err_mag_euc_Ftot_in_vs_euc_Ftot_out)
sd_err_mag_euc_Ftot_in_vs_euc_Ftot_out = sd (err_mag_euc_Ftot_in_vs_euc_Ftot_out)
median_err_mag_euc_Ftot_in_vs_euc_Ftot_out = median (err_mag_euc_Ftot_in_vs_euc_Ftot_out)
mad_err_mag_euc_Ftot_in_vs_euc_Ftot_out = mad (err_mag_euc_Ftot_in_vs_euc_Ftot_out)
cat ("\n\n", rs_name, " - err_mag_euc_Ftot_in_vs_euc_Ftot_out statistics:",
"\n\n mean = ", mean_err_mag_euc_Ftot_in_vs_euc_Ftot_out,
"\n sd = ", sd_err_mag_euc_Ftot_in_vs_euc_Ftot_out,
"\n\n median = ", median_err_mag_euc_Ftot_in_vs_euc_Ftot_out,
"\n mad = ", mad_err_mag_euc_Ftot_in_vs_euc_Ftot_out,
"\n", sep='')
#===============================================================================
#===============================================================================
#===============================================================================
#library (tidyverse)
rs_name = "Gurobi"
ref_y = 0
infile_02 = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_02_err_amt/default_runset/Gurobi.csv"
msa_dt_02 = read.csv (infile_02, header=TRUE, stringsAsFactors = FALSE)
infile_05 = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_05_err_amt/default_runset/Gurobi.csv"
msa_dt_05 = read.csv (infile_05, header=TRUE, stringsAsFactors = FALSE)
#infile_075 = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_075_err_amt/default_runset/Gurobi.csv"
#msa_dt_075 = read.csv (infile_075, header=TRUE, stringsAsFactors = FALSE)
infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_075_err_amt/default_runset/Gurobi.csv"
msa_dt = read.csv (infile, header=TRUE, stringsAsFactors = FALSE)
infile_10 = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/Gurobi.csv"
msa_dt_10 = read.csv (infile_10, header=TRUE, stringsAsFactors = FALSE)
infile_15 = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_15_err_amt/default_runset/Gurobi.csv"
msa_dt_15 = read.csv (infile_15, header=TRUE, stringsAsFactors = FALSE)
msa_dt = rbind (msa_dt_02, msa_dt_05, msa_dt_075)
msa_tib = as.tibble (msa_dt)
sorted_msa_tib = arrange (msa_tib, rsp_combined_err_label)
# Now that they are sorted by the 10 combined error labels,
# give an index to each rsrun within each of the combined error labels
# so that they can be spread out by index across a single facet plot.
# The index has no meaning other than to give it a unique identifier
# to spread along the x axis within each facet.
#sorted_msa_tib$idx = rep (1:600, 10) #1:dim(sorted_msa_tib)[1] # For the single set of 100 runs
sorted_msa_tib$idx = rep (1:200, 10) #1:dim(sorted_msa_tib)[1] # For the single set of 100 runs
msa_tib$rsp_combined_err_label [msa_tib$rsp_original_FP_const_rate == 0.05]
#===============================================================================
#===============================================================================
#===============================================================================
# 1 set of 20 outputs from 100 gen_20_basic_variants() runs
# on nectar and copied to glass.
#===============================================================================
#library (tidyverse)
#library (viridis)
rs_name = "Gurobi"
infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/Gurobi.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/Marxan_SA.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/SR_Forward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/SR_Backward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/ZL_Backward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/ZL_Forward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/UR_Forward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/UR_Backward.csv"
ref_y = 0.10
msa_dt = read.csv (infile, header=TRUE, stringsAsFactors = FALSE)
#===============================================================================
#===============================================================================
#===============================================================================
msa_tib = as.tibble (msa_dt)
sorted_msa_tib = arrange (msa_tib, rsp_combined_err_label)
# Now that they are sorted by the 10 combined error labels,
# give an index to each rsrun within each of the combined error labels
# so that they can be spread out by index across a single facet plot.
# The index has no meaning other than to give it a unique identifier
# to spread along the x axis within each facet.
sorted_msa_tib$idx = rep (1:200, 10) #1:dim(sorted_msa_tib)[1] # For the single set of 100 runs
#===============================================================================
#===============================================================================
#===============================================================================
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
#scale_color_viridis(discrete=TRUE) +
scale_color_manual(breaks = c("Base", "Wrap"), values=c("red", "blue")) +
ggtitle (paste0 (rs_name, " - Total input error vs. Total output error")) +
theme(plot.title = element_text(hjust = 0.5)) + # To center the title
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
#scale_color_viridis(discrete=TRUE) +
#scale_color_brewer(palette="Set1") +
scale_color_manual(breaks = c("Base", "Wrap"), values=c("red", "blue")) +
ggtitle (paste0 (rs_name, " - Total input error vs. Total output error by Error Class")) +
theme(plot.title = element_text(hjust = 0.5)) + # To center the title
facet_wrap (~ rsp_combined_err_label, nrow = 5) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
#----------
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac, y = err_mag_euc_Ftot_in_vs_euc_Ftot_out,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
# The graph from the previous call is bizarre and uninformative because
# you get some huge magnifications (up to 250) in the small input values.
# Here's a bit of code to select just the examples with magnification
# of 5 or less.
# Might be good to select for > 0 as well, or just remove the COR problems
# since magnification of a base of 0 is odd.
# Then again, maybe I need to redefine magnification there in some way
# because sometimes you DO get output error even without any input error.
# Not sure what the appropriate value would be there. Maybe magnification
# is the wrong idea there. Maybe I should just separate outputs for COR
# into a separate plot that just shows the error itself and then
# do magnification plots for APP values only.
# One other issue is that the input euc error is based on Ftot, which
# implicitly gives a different answer based on the number of coordinates
# that go into the formula, e.g., if you use (2) vs (2,2) vs (2,2,2) gives
# you 2, 2.828427, 3.464102. Need to figure out how to make sense of
# magnification in that context. You can't just replace Ftot with the
# euc combination of FP and FN because they're each weighted by the
# number of "pixels" involved. Maybe I can use something from how Ftot
# is calculated.
# Need to figure out how to do this with a filter() call since ggplot()
# doesn't work the way that I've subsetted for just the small mag values.
# For the moment, just doing a quick plot() call instead of ggplot(),
# just to get a sense of what it looks like.
indices_of_mags_le_5 = which (err_mag_euc_Ftot_in_vs_euc_Ftot_out <= 5)
in_err = sorted_msa_tib$rsp_euc_realized_Ftot_and_cost_in_err_frac [indices_of_mags_le_5]
mag = err_mag_euc_Ftot_in_vs_euc_Ftot_out [indices_of_mags_le_5]
plot (in_err, mag)
# This ggplot call gives the following error:
# Warning: Ignoring unknown aesthetics: NA
# Error: Aesthetics must be either length 1 or the same as the data (2000): x, y, colour
# ggplot (data = sorted_msa_tib) +
# geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac [indices_of_mags_le_5],
# y = err_mag_euc_Ftot_in_vs_euc_Ftot_out [indices_of_mags_le_5],
# # color = rsp_combined_err_label)) +
# color = rsp_base_wrap_str) [indices_of_mags_le_5]) +
# geom_hline (yintercept = ref_y, linetype="dashed",
# color = "black", size=0.5) +
# geom_abline (intercept=0, slope=1 #, linetype, color, size
# )
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
# FP and cost in err vs tot out err
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_FP_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
# FN and cost in err vs tot out err
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_FN_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
# median abs cost err vs tot out err
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_realized_median_abs_cost_err_frac,
y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
# Need to draw the 1:1 line because the plot
# is deceptive right now with roughly the
# same digits appearing on both axes, but
# the y values are 10 TIMES the x values.
# The 1:1 line is almost wholly below the
# dashed reference line, so the errors are
# quite big.
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
#----------------------------------------
# Make 2 plots, 1 for Base and 1 for Wrap.
# Put the index of the base problem (from 1 to 100) along the x axis.
# At each index on each plot, plot all 10 y values for that set of variants.
# index vs tot out err colored by Err Label, facetted by Base/Wrap side by side
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac,
color = rsp_combined_err_label)) +
#scale_color_viridis(discrete=TRUE) +
facet_wrap (~ rsp_base_wrap_str) +
# facet_wrap (~ rsp_base_wrap_str, nrow = 2) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5)
# index vs tot out colored by Err Label, facetted by Base/Wrap one above the other
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac,
color = rsp_combined_err_label)) +
#scale_color_viridis(discrete=TRUE) +
# facet_wrap (~ rsp_base_wrap_str) +
facet_wrap (~ rsp_base_wrap_str, nrow = 2) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5)
#----------------------------------------
# ggplot (data = sorted_msa_tib) +
# geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac,
# color = rsp_base_wrap_str)) +
# geom_hline (yintercept = ref_y, linetype="dashed",
# color = "black", size=0.5)
# index vs tot out err facetted by Combined Err Label, colored by Base/Wrap
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac,
color = rsp_base_wrap_str)) +
facet_wrap (~ rsp_combined_err_label, nrow = 5) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5)
# ggplot (data = sorted_msa_tib) +
# geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac,
# color = rsp_base_wrap_str)) +
# facet_wrap (~ rsp_combined_err_label, nrow = 3) +
# geom_hline (yintercept = ref_y, linetype="dashed",
# color = "black", size=0.5)
# index vs cost out err colored by Base/Wrap, facetted by Combined Err Label
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = rs_solution_cost_err_frac,
color = rsp_base_wrap_str)) +
facet_wrap (~ rsp_combined_err_label, nrow = 5) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_hline (yintercept = -ref_y, linetype="dashed",
color = "black", size=0.5)
# index vs abs cost out err colored by Base/Wrap, facetted by Combined Err Label
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = abs_rs_solution_cost_err_frac,
color = rsp_base_wrap_str)) +
facet_wrap (~ rsp_combined_err_label, nrow = 5)
# index vs rep shortfall err colored by Base/Wrap, facetted by Combined Err Label
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = rsr_APP_spp_rep_shortfall,
color = rsp_base_wrap_str)) +
facet_wrap (~ rsp_combined_err_label, nrow = 5) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5)
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
# Fooling around with the 20 outputs from 1 gen_20_basic_variants() run.
#===============================================================================
# Some relevant result column names to remember for use in the plots below.
# rsp_realized_FP_rate
# rsp_realized_FN_rate
# rsp_realized_Ftot_rate
# rsp_euc_realized_FP_and_cost_in_err_frac
# rsp_euc_realized_FN_and_cost_in_err_frac
# rsp_euc_realized_Ftot_and_cost_in_err_frac
#
# rsr_COR_euc_out_err_frac
# rsp_realized_median_abs_cost_err_frac
# rsp_realized_mean_abs_cost_err_frac
# rsp_realized_sd_abs_cost_err_frac
#library (tidyverse)
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_but_gurobi_1_easy_base/default_runset/18_easy.completedTzarEmulation/SR_Forward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_but_gurobi_1_easy_base/default_runset/19_easy.completedTzarEmulation/Marxan_SA.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_but_gurobi_1_easy_base/default_runset/24_easy.completedTzarEmulation/ZL_Backward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_but_gurobi_1_easy_base/default_runset/25_easy.completedTzarEmulation/Marxan_SA.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_but_gurobi_1_easy_base/default_runset/39_easy.completedTzarEmulation/Marxan_SA.csv"
infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_05_err_amt/default_runset/Gurobi.csv"
msa_dt = read.csv (infile, header=TRUE, stringsAsFactors = FALSE)
msa_tib = as.tibble (msa_dt)
sorted_msa_tib = arrange (msa_tib, rsp_combined_err_label)
sorted_msa_tib$idx = 1:20 #1:dim(sorted_msa_tib)[1]
# Need to find and load this value from somewhere in the results file,
# but I can't remember how it's labelled there at the moment, so
# I'm loading it by hand for now.
ref_y = 0.05
#ref_y = 0.02
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_realized_median_abs_cost_err_frac,
y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
# Need to draw the 1:1 line because the plot
# is deceptive right now with roughly the
# same digits appearing on both axes, but
# the y values are 10 TIMES the x values.
# The 1:1 line is almost wholly below the
# dashed reference line, so the errors are
# quite big.
geom_abline (intercept=0, slope=1 #, linetype, color, size
)
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
)
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_FP_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
)
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_FN_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
)
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
work_tib = select (msa_tib, rsr_UUID, rsp_UUID, rsp_cor_or_app_str, rsp_base_wrap_str, rsp_combined_err_label, rsp_file_name_prefix, rsp_prob_is_ok, rsp_num_PUs, rsp_num_spp, rsp_num_spp_per_PU, rsp_cost_error_bound, rsp_original_FP_const_rate, rsp_original_FN_const_rate, rsp_match_error_counts, rsp_FP_const_rate, rsp_FN_const_rate, rsp_realized_FP_rate, rsp_realized_FN_rate, rsp_realized_Ftot_rate, rsp_app_num_spp, rsp_app_num_PUs, cor_optimum_cost, rs_solution_cost, rs_solution_cost_err_frac, abs_rs_solution_cost_err_frac, rs_over_opt_cost_err_frac_of_possible_overcost, rs_under_opt_cost_err_frac_of_possible_undercost, rsr_COR_euc_out_err_frac, rsr_APP_spp_rep_shortfall, rsr_APP_solution_NUM_spp_covered, rsr_APP_solution_FRAC_spp_covered, RS_user_time, RS_system_time, RS_elapsed_time, RS_user_child_time, RS_sys_child_time)
err_tib = select (work_tib, rsr_UUID, rsp_UUID, rsp_cor_or_app_str, rsp_base_wrap_str, rsp_combined_err_label, rsp_cost_error_bound, rsp_original_FP_const_rate, rsp_original_FN_const_rate, rsp_match_error_counts, rsp_FP_const_rate, rsp_FN_const_rate, rsp_realized_FP_rate, rsp_realized_FN_rate, rsp_realized_Ftot_rate, rs_solution_cost_err_frac, abs_rs_solution_cost_err_frac, rs_over_opt_cost_err_frac_of_possible_overcost, rs_under_opt_cost_err_frac_of_possible_undercost, rsr_COR_euc_out_err_frac, rsr_APP_spp_rep_shortfall, rsr_APP_solution_NUM_spp_covered, rsr_APP_solution_FRAC_spp_covered, RS_user_time)
err_tib = arrange (err_tib, rsp_combined_err_label)
### Create an index value to use in making plots with a default x axis
idx = 1:dim(err_tib)[1]
err_tib$idx = idx
### Plot the absolute value of the cost error fraction against the apparent fraction of species covered.
ggplot(data=err_tib) + geom_point(mapping = aes(x = abs_rs_solution_cost_err_frac, y = rsr_APP_solution_FRAC_spp_covered), color = "blue")
### Facets
#ggplot (data = err_tib) +
ggplot (data = msa_tib) +
geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac))
#, color = rsp_base_wrap_str)
#ggplot (data = err_tib) +
# geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac), color = #rsp_base_wrap_str) +
# facet_wrap (~ rsp_combined_err_label, nrow = 10)
}
#===============================================================================
langfob/bdpg documentation built on Dec. 8, 2022, 5:33 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions dummy
#===============================================================================
#
# ggplot_20_variant_results.R
#
# This function is not meant to be called.
# It's only here as a dummy wrapper to keep R from executing this code when
# it builds the package.
# The code here is currently just a bunch of hacking around to do learn
# something about plotting with ggplot and at the same time, look at some
# very preliminary bdpg outputs from getting things working on nectar.
# It will probably evolve into real plotting code eventually, but right now,
# it's just a graveyard for code as I explore.
#
#===============================================================================
load_input_data <- function (rs_name, base_path, suffix,
num_sets = 4,
num_variants_per_set = 20,
num_combined_err_labels = 10)
{
infile = paste0 (base_path, rs_name, suffix)
msa_dt = read.csv (infile, header=TRUE, stringsAsFactors = FALSE)
msa_tib = as.tibble (msa_dt)
sorted_msa_tib = arrange (msa_tib, rsp_combined_err_label)
# Now that they are sorted by the 10 combined error labels,
# give an index to each rsrun within each of the combined error labels
# so that they can be spread out by index across a single facet plot.
# The index has no meaning other than to give it a unique identifier
# to spread along the x axis within each facet.
num_indices_per_combined_err_label =
num_sets * num_variants_per_set * num_combined_err_labels
sorted_msa_tib$idx = rep (1:num_indices_per_combined_err_label,
num_combined_err_labels)
unique_out_errs = unique (sorted_msa_tib$rsr_COR_euc_out_err_frac)
length (unique_out_errs)
sorted_unique_out_errs = sort (unique_out_errs)
cat ("\n\nhead (sorted_unique_OUT_errs) = \n")
print (head (sorted_unique_out_errs))
cat ("\ntail (sorted_unique_OUT_errs) = \n")
print (tail (sorted_unique_out_errs))
cat ("\n-----------------------")
unique_in_errs = unique (sorted_msa_tib$rsp_euc_realized_Ftot_and_cost_in_err_frac)
length (unique_in_errs)
sorted_unique_in_errs = sort (unique_in_errs)
cat ("\n\nhead (sorted_unique_IN_errs) = \n")
print (head (sorted_unique_in_errs))
cat ("\ntail (sorted_unique_IN_errs) = \n")
print (tail (sorted_unique_in_errs))
cat ("\n-----------------------")
return (sorted_msa_tib)
}
#===============================================================================
#' @importFrom stats mad quantile sd
compute_and_print_na_inf_stats <- function (rs_name,
out_err,
inp_err,
err_mag,
print_heading_str)
{
cat ("\n\n------------------------------------------------")
cat (paste0 ("\n\n----- ", print_heading_str, " -----"))
#----------
num_inp_na = length (which (is.na (inp_err)))
cat ("\n\nnum_inp_na for err_mag = ", num_inp_na)
num_inp_inf = length (which (is.infinite (inp_err)))
cat ("\nnum_inp_inf for err_mag = ", num_inp_inf)
num_inp_0 = length (which (inp_err == 0))
cat ("\nnum_inp_0 for err_mag = ", num_inp_0, "\n\n")
#----------
num_out_na = length (which (is.na (out_err)))
cat ("\n\nnum_out_na for err_mag = ", num_out_na)
num_out_inf = length (which (is.infinite (out_err)))
cat ("\nnum_out_inf for err_mag = ", num_out_inf)
num_out_0 = length (which (inp_err == 0))
cat ("\nnum_out_0 for err_mag = ", num_out_0, "\n\n")
#----------
num_mag_na = length (which (is.na (err_mag)))
cat ("\n\nnum_mag_na for err_mag = ", num_mag_na)
num_mag_inf = length (which (is.infinite (err_mag)))
cat ("\nnum_mag_inf for err_mag = ", num_mag_inf)
num_mag_0 = length (which (inp_err == 0))
cat ("\nnum_mag_0 for err_mag = ", num_mag_0, "\n\n")
#----------
# Set infinite values to NA.
err_mag [is.infinite (err_mag)] = NA
mean_err_mag = mean (err_mag, na.rm = TRUE)
sd_err_mag = sd (err_mag, na.rm = TRUE)
median_err_mag = median (err_mag, na.rm = TRUE)
mad_err_mag = mad (err_mag, na.rm = TRUE)
cat (rs_name, " - Error magnification statistics (i.e., outErr/inErr):",
"\n\n mean = ", mean_err_mag,
"\n sd = ", sd_err_mag,
"\n\n median = ", median_err_mag,
"\n mad = ", mad_err_mag,
"\n\n", sep='')
cat ("Quantiles:\n")
print (quantile (err_mag, na.rm = TRUE))
#browser()
}
compute_and_print_mag_stats <- function (rs_name,
out_err,
inp_err
, rsr_UUID
, rsp_combined_err_label
)
{
err_mag = out_err / inp_err
inf_locs = which (is.infinite (err_mag))
if (length (inf_locs) > 0)
{
for (iii in 1:length (inf_locs))
{
cat ("\nInf err_mag [", inf_locs[iii],
"], rsr_UUID = ", rsr_UUID [inf_locs[iii]], ", ", rsp_combined_err_label [inf_locs[iii]], ": inp_err = ", inp_err[inf_locs[iii]], ",
out_err = ", out_err[inf_locs[iii]])
}
}
compute_and_print_na_inf_stats (rs_name,
out_err,
inp_err,
err_mag,
"FOR ALL DATA")
# cat ("\n\n------------------------------------------------")
# cat ("\n\n----- FOR ALL DATA -----")
# num_mag_na = length (which (is.na (err_mag)))
# cat ("\n\nnum_mag_na for err_mag = ", num_mag_na)
# num_inf = length (which (is.infinite (err_mag)))
# cat ("\nnum_inf for err_mag = ", num_inf, "\n\n")
# err_mag [is.infinite (err_mag)] = NA
#
# mean_err_mag = mean (err_mag, na.rm = TRUE)
# sd_err_mag = sd (err_mag, na.rm = TRUE)
#
# median_err_mag = median (err_mag, na.rm = TRUE)
# mad_err_mag = mad (err_mag, na.rm = TRUE)
#
# cat (rs_name, " - Error magnification statistics (i.e., outErr/inErr):",
# "\n\n mean = ", mean_err_mag,
# "\n sd = ", sd_err_mag,
# "\n\n median = ", median_err_mag,
# "\n mad = ", mad_err_mag,
# "\n\n", sep='')
#
# cat ("Quantiles:\n")
# print (quantile (err_mag, na.rm = TRUE))
#--------------------
# cat ("\n\n------------------------------------------------")
# cat ("\n\n----- FOR input error >= 0.01 -----")
ge1_indices = which (inp_err >= 0.01)
num_indices_ge1 = length (ge1_indices)
cat ("\n\nnum_indices_ge1 = ", num_indices_ge1)
err_mag = err_mag [ge1_indices]
compute_and_print_na_inf_stats (rs_name,
out_err,
inp_err,
err_mag,
"FOR input error >= 0.01")
# num_mag_na = length (which (is.na (err_mag)))
# cat ("\n\nnum_mag_na for err_mag = ", num_mag_na)
#
# num_inf = length (which (is.infinite (err_mag)))
# cat ("\nnum_inf for err_mag = ", num_inf, "\n\n")
# err_mag [is.infinite (err_mag)] = NA
#
# mean_err_mag = mean (err_mag, na.rm = TRUE)
# sd_err_mag = sd (err_mag, na.rm = TRUE)
#
# median_err_mag = median (err_mag, na.rm = TRUE)
# mad_err_mag = mad (err_mag, na.rm = TRUE)
#
# cat (rs_name, " - Error magnification statistics (i.e., outErr/inErr):",
# "\n\n mean = ", mean_err_mag,
# "\n sd = ", sd_err_mag,
# "\n\n median = ", median_err_mag,
# "\n mad = ", mad_err_mag,
# "\n\n", sep='')
#
# cat ("Quantiles:\n")
# print (quantile (err_mag, na.rm = TRUE))
}
#===============================================================================
gg_multiple_err_amts <- function (rs_name, base_path, suffix,
num_sets = 4,
num_variants_per_set = 20,
num_combined_err_labels = 10)
{
sorted_msa_tib = load_input_data (rs_name, base_path, suffix)
compute_and_print_mag_stats (rs_name,
sorted_msa_tib$rsr_COR_euc_out_err_frac,
sorted_msa_tib$rsp_euc_realized_Ftot_and_cost_in_err_frac
, sorted_msa_tib$rsr_UUID
, sorted_msa_tib$rsp_combined_err_label)
return (sorted_msa_tib)
}
#===============================================================================
gg_eucInTot_vs_eucOutTot_all_on_1 <- function (rs_name,
sorted_msa_tib,
ref_y = 0)
{
ggplot (data = sorted_msa_tib) +
aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac,
y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str) +
geom_point() +
scale_color_manual(breaks = c("Base", "Wrap"), values=c("red", "blue")) +
stat_smooth(method = "lm", col = "green") +
# ggplot (data = sorted_msa_tib) +
# geom_point (mapping = ) +
# #scale_color_viridis(discrete=TRUE) +
# scale_color_manual(breaks = c("Base", "Wrap"), values=c("red", "blue")) +
#
ggtitle (paste0 (rs_name, " - Total input error vs. Total output error")) +
theme(plot.title = element_text(hjust = 0.5)) + # To center the title
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
}
#===============================================================================
gg_eucInTot_vs_eucOutTot_facetted_by_err_label <- function (rs_name,
sorted_msa_tib,
ref_y = 0)
{
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac,
y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
#scale_color_viridis(discrete=TRUE) +
#scale_color_brewer(palette="Set1") +
scale_color_manual(breaks = c("Base", "Wrap"), values=c("red", "blue")) +
ggtitle (paste0 (rs_name, " - Total input error vs. Total output error by Error Class")) +
theme(plot.title = element_text(hjust = 0.5)) + # To center the title
facet_wrap (~ rsp_combined_err_label, nrow = 5) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
}
#===============================================================================
dummy <- function()
{
#library (tidyverse)
base_path = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/"
suffix = ".combined_results.csv"
rs_name = "Gurobi"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "Marxan_SA"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "ZL_Backward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "ZL_Forward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "UR_Backward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "UR_Forward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "SR_Backward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
rs_name = "SR_Forward"
sorted_msa_tib = gg_multiple_err_amts (rs_name, base_path, suffix)
gg_eucInTot_vs_eucOutTot_all_on_1 (rs_name, sorted_msa_tib, ref_y = 0)
gg_eucInTot_vs_eucOutTot_facetted_by_err_label (rs_name, sorted_msa_tib, ref_y = 0)
}
#===============================================================================
#===============================================================================
#===============================================================================
gg_20 <- function ()
{
#===============================================================================
#===============================================================================
#===============================================================================
# 4 sets of 20 outputs from 100 gen_20_basic_variants() runs
# on nectar and copied to glass.
#===============================================================================
### Load necessary libraries.
#library (tidyverse)
# rs_name = "Gurobi"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/Gurobi.combined_results.csv"
infile = paste0 (base_path, rs_name, suffix)
# rs_name = "Marxan_SA"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/Marxan_SA.combined_results.csv"
# rs_name = "ZL_Backward"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/ZL_Backward.combined_results.csv"
# rs_name = "ZL_Forward"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/ZL_Forward.combined_results.csv"
# rs_name = "SR_Forward"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/SR_Forward.combined_results.csv"
# rs_name = "UR_Forward"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/UR_Forward.combined_results.csv"
# rs_name = "UR_Backward"
# infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/UR_Backward.combined_results.csv"
rs_name = "SR_Backward"
infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_Combined_err_amts/SR_Backward.combined_results.csv"
ref_y = 0
msa_dt = read.csv (infile, header=TRUE, stringsAsFactors = FALSE)
msa_tib = as.tibble (msa_dt)
sorted_msa_tib = arrange (msa_tib, rsp_combined_err_label)
# Now that they are sorted by the 10 combined error labels,
# give an index to each rsrun within each of the combined error labels
# so that they can be spread out by index across a single facet plot.
# The index has no meaning other than to give it a unique identifier
# to spread along the x axis within each facet.
sorted_msa_tib$idx = rep (1:800, 10) #1:dim(sorted_msa_tib)[1] # For the 4 combined sets 0f 100 runs each
#----------
# NOTE: the result of selecting for NA gives 201 instead of 200 results.
# I would expect 200 since there are 200 correct values that have no
# input error. Not sure what 1 extra NA is from.
err_mag_euc_Ftot_in_vs_euc_Ftot_out =
sorted_msa_tib$rsr_COR_euc_out_err_frac /
sorted_msa_tib$rsp_euc_realized_Ftot_and_cost_in_err_frac
err_mag_euc_Ftot_in_vs_euc_Ftot_out [is.na (err_mag_euc_Ftot_in_vs_euc_Ftot_out)] = 0
mean_err_mag_euc_Ftot_in_vs_euc_Ftot_out = mean (err_mag_euc_Ftot_in_vs_euc_Ftot_out)
sd_err_mag_euc_Ftot_in_vs_euc_Ftot_out = sd (err_mag_euc_Ftot_in_vs_euc_Ftot_out)
median_err_mag_euc_Ftot_in_vs_euc_Ftot_out = median (err_mag_euc_Ftot_in_vs_euc_Ftot_out)
mad_err_mag_euc_Ftot_in_vs_euc_Ftot_out = mad (err_mag_euc_Ftot_in_vs_euc_Ftot_out)
cat ("\n\n", rs_name, " - err_mag_euc_Ftot_in_vs_euc_Ftot_out statistics:",
"\n\n mean = ", mean_err_mag_euc_Ftot_in_vs_euc_Ftot_out,
"\n sd = ", sd_err_mag_euc_Ftot_in_vs_euc_Ftot_out,
"\n\n median = ", median_err_mag_euc_Ftot_in_vs_euc_Ftot_out,
"\n mad = ", mad_err_mag_euc_Ftot_in_vs_euc_Ftot_out,
"\n", sep='')
#===============================================================================
#===============================================================================
#===============================================================================
#library (tidyverse)
rs_name = "Gurobi"
ref_y = 0
infile_02 = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_02_err_amt/default_runset/Gurobi.csv"
msa_dt_02 = read.csv (infile_02, header=TRUE, stringsAsFactors = FALSE)
infile_05 = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_05_err_amt/default_runset/Gurobi.csv"
msa_dt_05 = read.csv (infile_05, header=TRUE, stringsAsFactors = FALSE)
#infile_075 = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_075_err_amt/default_runset/Gurobi.csv"
#msa_dt_075 = read.csv (infile_075, header=TRUE, stringsAsFactors = FALSE)
infile = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_075_err_amt/default_runset/Gurobi.csv"
msa_dt = read.csv (infile, header=TRUE, stringsAsFactors = FALSE)
infile_10 = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/Gurobi.csv"
msa_dt_10 = read.csv (infile_10, header=TRUE, stringsAsFactors = FALSE)
infile_15 = "/Users/bill/D/Projects/ProblemDifficulty/Results/bdpg_20_variants_all_rs_easy_base/bdpg_20_variants_all_rs_easy_base_15_err_amt/default_runset/Gurobi.csv"
msa_dt_15 = read.csv (infile_15, header=TRUE, stringsAsFactors = FALSE)
msa_dt = rbind (msa_dt_02, msa_dt_05, msa_dt_075)
msa_tib = as.tibble (msa_dt)
sorted_msa_tib = arrange (msa_tib, rsp_combined_err_label)
# Now that they are sorted by the 10 combined error labels,
# give an index to each rsrun within each of the combined error labels
# so that they can be spread out by index across a single facet plot.
# The index has no meaning other than to give it a unique identifier
# to spread along the x axis within each facet.
#sorted_msa_tib$idx = rep (1:600, 10) #1:dim(sorted_msa_tib)[1] # For the single set of 100 runs
sorted_msa_tib$idx = rep (1:200, 10) #1:dim(sorted_msa_tib)[1] # For the single set of 100 runs
msa_tib$rsp_combined_err_label [msa_tib$rsp_original_FP_const_rate == 0.05]
#===============================================================================
#===============================================================================
#===============================================================================
# 1 set of 20 outputs from 100 gen_20_basic_variants() runs
# on nectar and copied to glass.
#===============================================================================
#library (tidyverse)
#library (viridis)
rs_name = "Gurobi"
infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/Gurobi.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/Marxan_SA.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/SR_Forward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/SR_Backward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/ZL_Backward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/ZL_Forward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/UR_Forward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_10_err_amt/default_runset/UR_Backward.csv"
ref_y = 0.10
msa_dt = read.csv (infile, header=TRUE, stringsAsFactors = FALSE)
#===============================================================================
#===============================================================================
#===============================================================================
msa_tib = as.tibble (msa_dt)
sorted_msa_tib = arrange (msa_tib, rsp_combined_err_label)
# Now that they are sorted by the 10 combined error labels,
# give an index to each rsrun within each of the combined error labels
# so that they can be spread out by index across a single facet plot.
# The index has no meaning other than to give it a unique identifier
# to spread along the x axis within each facet.
sorted_msa_tib$idx = rep (1:200, 10) #1:dim(sorted_msa_tib)[1] # For the single set of 100 runs
#===============================================================================
#===============================================================================
#===============================================================================
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
#scale_color_viridis(discrete=TRUE) +
scale_color_manual(breaks = c("Base", "Wrap"), values=c("red", "blue")) +
ggtitle (paste0 (rs_name, " - Total input error vs. Total output error")) +
theme(plot.title = element_text(hjust = 0.5)) + # To center the title
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
#scale_color_viridis(discrete=TRUE) +
#scale_color_brewer(palette="Set1") +
scale_color_manual(breaks = c("Base", "Wrap"), values=c("red", "blue")) +
ggtitle (paste0 (rs_name, " - Total input error vs. Total output error by Error Class")) +
theme(plot.title = element_text(hjust = 0.5)) + # To center the title
facet_wrap (~ rsp_combined_err_label, nrow = 5) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
#----------
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac, y = err_mag_euc_Ftot_in_vs_euc_Ftot_out,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
# The graph from the previous call is bizarre and uninformative because
# you get some huge magnifications (up to 250) in the small input values.
# Here's a bit of code to select just the examples with magnification
# of 5 or less.
# Might be good to select for > 0 as well, or just remove the COR problems
# since magnification of a base of 0 is odd.
# Then again, maybe I need to redefine magnification there in some way
# because sometimes you DO get output error even without any input error.
# Not sure what the appropriate value would be there. Maybe magnification
# is the wrong idea there. Maybe I should just separate outputs for COR
# into a separate plot that just shows the error itself and then
# do magnification plots for APP values only.
# One other issue is that the input euc error is based on Ftot, which
# implicitly gives a different answer based on the number of coordinates
# that go into the formula, e.g., if you use (2) vs (2,2) vs (2,2,2) gives
# you 2, 2.828427, 3.464102. Need to figure out how to make sense of
# magnification in that context. You can't just replace Ftot with the
# euc combination of FP and FN because they're each weighted by the
# number of "pixels" involved. Maybe I can use something from how Ftot
# is calculated.
# Need to figure out how to do this with a filter() call since ggplot()
# doesn't work the way that I've subsetted for just the small mag values.
# For the moment, just doing a quick plot() call instead of ggplot(),
# just to get a sense of what it looks like.
indices_of_mags_le_5 = which (err_mag_euc_Ftot_in_vs_euc_Ftot_out <= 5)
in_err = sorted_msa_tib$rsp_euc_realized_Ftot_and_cost_in_err_frac [indices_of_mags_le_5]
mag = err_mag_euc_Ftot_in_vs_euc_Ftot_out [indices_of_mags_le_5]
plot (in_err, mag)
# This ggplot call gives the following error:
# Warning: Ignoring unknown aesthetics: NA
# Error: Aesthetics must be either length 1 or the same as the data (2000): x, y, colour
# ggplot (data = sorted_msa_tib) +
# geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac [indices_of_mags_le_5],
# y = err_mag_euc_Ftot_in_vs_euc_Ftot_out [indices_of_mags_le_5],
# # color = rsp_combined_err_label)) +
# color = rsp_base_wrap_str) [indices_of_mags_le_5]) +
# geom_hline (yintercept = ref_y, linetype="dashed",
# color = "black", size=0.5) +
# geom_abline (intercept=0, slope=1 #, linetype, color, size
# )
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
#----------------------------------------
# FP and cost in err vs tot out err
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_FP_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
# FN and cost in err vs tot out err
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_FN_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
# median abs cost err vs tot out err
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_realized_median_abs_cost_err_frac,
y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
# Need to draw the 1:1 line because the plot
# is deceptive right now with roughly the
# same digits appearing on both axes, but
# the y values are 10 TIMES the x values.
# The 1:1 line is almost wholly below the
# dashed reference line, so the errors are
# quite big.
geom_abline (intercept=0, slope=1 #, linetype, color, size
) +
geom_abline (intercept=0, slope=5 #, linetype, color, size
) +
geom_abline (intercept=0, slope=10 #, linetype, color, size
)
#----------------------------------------
# Make 2 plots, 1 for Base and 1 for Wrap.
# Put the index of the base problem (from 1 to 100) along the x axis.
# At each index on each plot, plot all 10 y values for that set of variants.
# index vs tot out err colored by Err Label, facetted by Base/Wrap side by side
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac,
color = rsp_combined_err_label)) +
#scale_color_viridis(discrete=TRUE) +
facet_wrap (~ rsp_base_wrap_str) +
# facet_wrap (~ rsp_base_wrap_str, nrow = 2) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5)
# index vs tot out colored by Err Label, facetted by Base/Wrap one above the other
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac,
color = rsp_combined_err_label)) +
#scale_color_viridis(discrete=TRUE) +
# facet_wrap (~ rsp_base_wrap_str) +
facet_wrap (~ rsp_base_wrap_str, nrow = 2) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5)
#----------------------------------------
# ggplot (data = sorted_msa_tib) +
# geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac,
# color = rsp_base_wrap_str)) +
# geom_hline (yintercept = ref_y, linetype="dashed",
# color = "black", size=0.5)
# index vs tot out err facetted by Combined Err Label, colored by Base/Wrap
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac,
color = rsp_base_wrap_str)) +
facet_wrap (~ rsp_combined_err_label, nrow = 5) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5)
# ggplot (data = sorted_msa_tib) +
# geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac,
# color = rsp_base_wrap_str)) +
# facet_wrap (~ rsp_combined_err_label, nrow = 3) +
# geom_hline (yintercept = ref_y, linetype="dashed",
# color = "black", size=0.5)
# index vs cost out err colored by Base/Wrap, facetted by Combined Err Label
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = rs_solution_cost_err_frac,
color = rsp_base_wrap_str)) +
facet_wrap (~ rsp_combined_err_label, nrow = 5) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_hline (yintercept = -ref_y, linetype="dashed",
color = "black", size=0.5)
# index vs abs cost out err colored by Base/Wrap, facetted by Combined Err Label
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = abs_rs_solution_cost_err_frac,
color = rsp_base_wrap_str)) +
facet_wrap (~ rsp_combined_err_label, nrow = 5)
# index vs rep shortfall err colored by Base/Wrap, facetted by Combined Err Label
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = idx, y = rsr_APP_spp_rep_shortfall,
color = rsp_base_wrap_str)) +
facet_wrap (~ rsp_combined_err_label, nrow = 5) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5)
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
# Fooling around with the 20 outputs from 1 gen_20_basic_variants() run.
#===============================================================================
# Some relevant result column names to remember for use in the plots below.
# rsp_realized_FP_rate
# rsp_realized_FN_rate
# rsp_realized_Ftot_rate
# rsp_euc_realized_FP_and_cost_in_err_frac
# rsp_euc_realized_FN_and_cost_in_err_frac
# rsp_euc_realized_Ftot_and_cost_in_err_frac
#
# rsr_COR_euc_out_err_frac
# rsp_realized_median_abs_cost_err_frac
# rsp_realized_mean_abs_cost_err_frac
# rsp_realized_sd_abs_cost_err_frac
#library (tidyverse)
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_but_gurobi_1_easy_base/default_runset/18_easy.completedTzarEmulation/SR_Forward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_but_gurobi_1_easy_base/default_runset/19_easy.completedTzarEmulation/Marxan_SA.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_but_gurobi_1_easy_base/default_runset/24_easy.completedTzarEmulation/ZL_Backward.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_but_gurobi_1_easy_base/default_runset/25_easy.completedTzarEmulation/Marxan_SA.csv"
#infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_but_gurobi_1_easy_base/default_runset/39_easy.completedTzarEmulation/Marxan_SA.csv"
infile = "/Users/bill/tzar/outputdata/bdpg_20_variants_all_rs_easy_base_05_err_amt/default_runset/Gurobi.csv"
msa_dt = read.csv (infile, header=TRUE, stringsAsFactors = FALSE)
msa_tib = as.tibble (msa_dt)
sorted_msa_tib = arrange (msa_tib, rsp_combined_err_label)
sorted_msa_tib$idx = 1:20 #1:dim(sorted_msa_tib)[1]
# Need to find and load this value from somewhere in the results file,
# but I can't remember how it's labelled there at the moment, so
# I'm loading it by hand for now.
ref_y = 0.05
#ref_y = 0.02
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_realized_median_abs_cost_err_frac,
y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
# Need to draw the 1:1 line because the plot
# is deceptive right now with roughly the
# same digits appearing on both axes, but
# the y values are 10 TIMES the x values.
# The 1:1 line is almost wholly below the
# dashed reference line, so the errors are
# quite big.
geom_abline (intercept=0, slope=1 #, linetype, color, size
)
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_Ftot_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
# color = rsp_combined_err_label)) +
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
)
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_FP_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
)
ggplot (data = sorted_msa_tib) +
geom_point (mapping = aes(x = rsp_euc_realized_FN_and_cost_in_err_frac, y = rsr_COR_euc_out_err_frac,
color = rsp_base_wrap_str)) +
geom_hline (yintercept = ref_y, linetype="dashed",
color = "black", size=0.5) +
geom_abline (intercept=0, slope=1 #, linetype, color, size
)
#===============================================================================
#===============================================================================
#===============================================================================
#===============================================================================
work_tib = select (msa_tib, rsr_UUID, rsp_UUID, rsp_cor_or_app_str, rsp_base_wrap_str, rsp_combined_err_label, rsp_file_name_prefix, rsp_prob_is_ok, rsp_num_PUs, rsp_num_spp, rsp_num_spp_per_PU, rsp_cost_error_bound, rsp_original_FP_const_rate, rsp_original_FN_const_rate, rsp_match_error_counts, rsp_FP_const_rate, rsp_FN_const_rate, rsp_realized_FP_rate, rsp_realized_FN_rate, rsp_realized_Ftot_rate, rsp_app_num_spp, rsp_app_num_PUs, cor_optimum_cost, rs_solution_cost, rs_solution_cost_err_frac, abs_rs_solution_cost_err_frac, rs_over_opt_cost_err_frac_of_possible_overcost, rs_under_opt_cost_err_frac_of_possible_undercost, rsr_COR_euc_out_err_frac, rsr_APP_spp_rep_shortfall, rsr_APP_solution_NUM_spp_covered, rsr_APP_solution_FRAC_spp_covered, RS_user_time, RS_system_time, RS_elapsed_time, RS_user_child_time, RS_sys_child_time)
err_tib = select (work_tib, rsr_UUID, rsp_UUID, rsp_cor_or_app_str, rsp_base_wrap_str, rsp_combined_err_label, rsp_cost_error_bound, rsp_original_FP_const_rate, rsp_original_FN_const_rate, rsp_match_error_counts, rsp_FP_const_rate, rsp_FN_const_rate, rsp_realized_FP_rate, rsp_realized_FN_rate, rsp_realized_Ftot_rate, rs_solution_cost_err_frac, abs_rs_solution_cost_err_frac, rs_over_opt_cost_err_frac_of_possible_overcost, rs_under_opt_cost_err_frac_of_possible_undercost, rsr_COR_euc_out_err_frac, rsr_APP_spp_rep_shortfall, rsr_APP_solution_NUM_spp_covered, rsr_APP_solution_FRAC_spp_covered, RS_user_time)
err_tib = arrange (err_tib, rsp_combined_err_label)
### Create an index value to use in making plots with a default x axis
idx = 1:dim(err_tib)[1]
err_tib$idx = idx
### Plot the absolute value of the cost error fraction against the apparent fraction of species covered.
ggplot(data=err_tib) + geom_point(mapping = aes(x = abs_rs_solution_cost_err_frac, y = rsr_APP_solution_FRAC_spp_covered), color = "blue")
### Facets
#ggplot (data = err_tib) +
ggplot (data = msa_tib) +
geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac))
#, color = rsp_base_wrap_str)
#ggplot (data = err_tib) +
# geom_point (mapping = aes(x = idx, y = rsr_COR_euc_out_err_frac), color = #rsp_base_wrap_str) +
# facet_wrap (~ rsp_combined_err_label, nrow = 10)
}
#===============================================================================
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.