## ----setup, include = FALSE---------------------------------------------------
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.width = 8,
  fig.height = 6,
  warning = FALSE,
  message = FALSE
)

## ----libraries----------------------------------------------------------------
library(searchAnalyzeR)
library(ggplot2)

# Load additional packages for enhanced visualizations
if (requireNamespace("patchwork", quietly = TRUE)) {
  library(patchwork)
}

## ----strategy_a---------------------------------------------------------------
strategy_A <- list(
  name = "Broad COVID Research Strategy",
  terms = c(
    "covid-19",
    "coronavirus", 
    "sars-cov-2",
    "pandemic",
    "covid"
  ),
  description = "Broad approach capturing general COVID-19 research across all domains",
  databases = c("PubMed"),
  date_range = as.Date(c("2020-01-01", "2024-12-31")),
  filters = list(
    language = "English",
    article_types = c("Journal Article", "Review", "Clinical Trial")
  ),
  search_date = Sys.time()
)

cat("Strategy A (Broad COVID Research):\n")
cat("Terms:", paste(strategy_A$terms, collapse = " OR "), "\n")
cat("Description:", strategy_A$description, "\n")

## ----strategy_b---------------------------------------------------------------
strategy_B <- list(
  name = "Targeted Clinical COVID Strategy", 
  terms = c(
    "covid-19 clinical trial",
    "covid-19 randomized controlled trial",
    "covid-19 systematic review",
    "covid-19 vaccine efficacy",
    "covid-19 treatment outcomes"
  ),
  description = "Targeted approach focusing on high-quality clinical evidence",
  databases = c("PubMed"),
  date_range = as.Date(c("2020-01-01", "2024-12-31")),
  filters = list(
    language = "English", 
    article_types = c("Journal Article", "Review", "Clinical Trial")
  ),
  search_date = Sys.time()
)

cat("Strategy B (Targeted Clinical COVID):\n")
cat("Terms:", paste(strategy_B$terms, collapse = " OR "), "\n")
cat("Description:", strategy_B$description, "\n")

## ----execute_searches---------------------------------------------------------
# Execute Strategy A
cat("=== EXECUTING STRATEGY A: Broad COVID Research ===\n")
results_A <- search_pubmed(
  search_terms = strategy_A$terms,
  max_results = 200,
  date_range = strategy_A$date_range
)

cat("Strategy A completed. Retrieved", nrow(results_A), "articles.\n\n")

# Execute Strategy B  
cat("=== EXECUTING STRATEGY B: Targeted Clinical COVID ===\n")
results_B <- search_pubmed(
  search_terms = strategy_B$terms,
  max_results = 200,
  date_range = strategy_B$date_range
)

cat("Strategy B completed. Retrieved", nrow(results_B), "articles.\n")

## ----process_data-------------------------------------------------------------
# Standardize both result sets
cat("Standardizing search results...\n")
standardized_A <- std_search_results(results_A, source_format = "pubmed")
standardized_B <- std_search_results(results_B, source_format = "pubmed")

# Add strategy identifiers
standardized_A$strategy <- "Broad_COVID_Research"
standardized_B$strategy <- "Targeted_Clinical_COVID"

# Detect duplicates within each strategy
dedup_A <- detect_dupes(standardized_A, method = "exact")
dedup_B <- detect_dupes(standardized_B, method = "exact")

# Report deduplication results
cat("Strategy A - Total:", nrow(dedup_A), 
    "Unique:", sum(!dedup_A$duplicate), 
    "Duplicates:", sum(dedup_A$duplicate), "\n")
cat("Strategy B - Total:", nrow(dedup_B), 
    "Unique:", sum(!dedup_B$duplicate), 
    "Duplicates:", sum(dedup_B$duplicate), "\n")

## ----create_gold_standard-----------------------------------------------------
# Filter out duplicates for analysis
unique_A <- dedup_A[!dedup_A$duplicate, ]
unique_B <- dedup_B[!dedup_B$duplicate, ]

# High-confidence terms that indicate quality COVID research
high_confidence_patterns <- c(
  "randomized", "controlled trial", "systematic review", "meta-analysis",
  "clinical trial", "vaccine efficacy", "treatment outcome", "placebo",
  "double-blind", "multicenter", "cohort study", "case-control"
)

# Function to count high-confidence patterns
count_patterns <- function(data, patterns) {
  combined_text <- tolower(paste(data$title, data$abstract, sep = " "))
  sapply(combined_text, function(text) {
    sum(sapply(patterns, function(pattern) grepl(pattern, text, fixed = TRUE)))
  })
}

# Articles that appear in both strategies (high confidence due to overlap)
overlap_ids <- intersect(unique_A$id, unique_B$id)

# Articles with multiple high-confidence patterns
pattern_counts_A <- count_patterns(unique_A, high_confidence_patterns)
pattern_counts_B <- count_patterns(unique_B, high_confidence_patterns)

multi_pattern_A <- unique_A$id[pattern_counts_A >= 2]
multi_pattern_B <- unique_B$id[pattern_counts_B >= 2]

# Articles with "systematic review" or "meta-analysis" in title
systematic_review_pattern <- "systematic review|meta-analysis"
systematic_A <- unique_A$id[grepl(systematic_review_pattern, tolower(unique_A$title))]
systematic_B <- unique_B$id[grepl(systematic_review_pattern, tolower(unique_B$title))]

# Combine for gold standard
gold_standard_ids <- unique(c(
  overlap_ids,      # High confidence: found by both strategies
  systematic_A,     # Very high confidence: systematic reviews from A
  systematic_B,     # Very high confidence: systematic reviews from B  
  multi_pattern_A,  # High confidence: multiple quality indicators from A
  multi_pattern_B   # High confidence: multiple quality indicators from B
))

cat("Gold standard created with", length(gold_standard_ids), "high-confidence relevant articles\n")
cat("- Overlap between strategies:", length(overlap_ids), "articles\n")
cat("- Systematic reviews Strategy A:", length(systematic_A), "articles\n")
cat("- Systematic reviews Strategy B:", length(systematic_B), "articles\n")

## ----performance_analysis-----------------------------------------------------
# Initialize analyzers for both strategies
unique_A_ids <- unique_A$id
unique_B_ids <- unique_B$id

analyzer_A <- SearchAnalyzer$new(
  search_results = unique_A,
  gold_standard = gold_standard_ids,
  search_strategy = strategy_A
)

analyzer_B <- SearchAnalyzer$new(
  search_results = unique_B,
  gold_standard = gold_standard_ids,
  search_strategy = strategy_B
)

# Calculate comprehensive metrics
metrics_A <- analyzer_A$calculate_metrics()
metrics_B <- analyzer_B$calculate_metrics()

# Display key metrics
cat("STRATEGY A (Broad COVID Research) PERFORMANCE:\n")
cat("Total Articles Retrieved:", nrow(unique_A), "\n")
cat("Precision:", round(metrics_A$precision_recall$precision, 3), "\n")
cat("Recall:", round(metrics_A$precision_recall$recall, 3), "\n")
cat("F1 Score:", round(metrics_A$precision_recall$f1_score, 3), "\n")

cat("\nSTRATEGY B (Targeted Clinical COVID) PERFORMANCE:\n")
cat("Total Articles Retrieved:", nrow(unique_B), "\n")
cat("Precision:", round(metrics_B$precision_recall$precision, 3), "\n")
cat("Recall:", round(metrics_B$precision_recall$recall, 3), "\n")
cat("F1 Score:", round(metrics_B$precision_recall$f1_score, 3), "\n")

## ----statistical_comparison---------------------------------------------------
# Compare strategies using McNemar's test
comparison_result <- compare_strategies(
  strategy1_results = unique_A_ids,
  strategy2_results = unique_B_ids,
  gold_standard = gold_standard_ids,
  test_type = "mcnemar"
)

cat("STATISTICAL COMPARISON RESULTS:\n")
cat("Test Used:", comparison_result$test, "\n")
cat("P-value:", round(comparison_result$p_value, 4), "\n")
cat("Statistically Significant:", comparison_result$significant, "\n")

if (!is.null(comparison_result$difference)) {
  cat("\nPERFORMANCE DIFFERENCES (B - A):\n")
  cat("Precision Difference:", round(comparison_result$difference$precision_diff, 3), "\n")
  cat("Recall Difference:", round(comparison_result$difference$recall_diff, 3), "\n")
  cat("F1 Score Difference:", round(comparison_result$difference$f1_diff, 3), "\n")
}

## ----complementarity_analysis-------------------------------------------------
# Calculate detailed strategy comparison metrics
strategy_comparison <- calc_strategy_comparison(
  strategy1_results = unique_A_ids,
  strategy2_results = unique_B_ids,
  gold_standard = gold_standard_ids
)

cat("ENHANCED OVERLAP ANALYSIS:\n")
cat("Total Unique Articles (Combined):", strategy_comparison$overlap_analysis$total_unique, "\n")
cat("Overlap Between Strategies:", strategy_comparison$overlap_analysis$overlap_count, "\n")
cat("Unique to Strategy A (Broad):", strategy_comparison$overlap_analysis$unique_to_strategy1, "\n")
cat("Unique to Strategy B (Targeted):", strategy_comparison$overlap_analysis$unique_to_strategy2, "\n")
cat("Overlap Percentage:", round(strategy_comparison$overlap_analysis$overlap_percentage, 1), "%\n")

cat("\nCOMPLEMENTARITY ANALYSIS:\n")
cat("Added Recall by Strategy A:", round(strategy_comparison$complementarity$added_recall_by_strategy1, 3), "\n")
cat("Added Recall by Strategy B:", round(strategy_comparison$complementarity$added_recall_by_strategy2, 3), "\n")
cat("Synergy Score:", round(strategy_comparison$complementarity$synergy_score, 3), "\n")

## ----temporal_analysis--------------------------------------------------------
# Calculate temporal coverage for both strategies
temporal_A <- calc_temporal_coverage(unique_A, target_date_range = strategy_A$date_range)
temporal_B <- calc_temporal_coverage(unique_B, target_date_range = strategy_B$date_range)

cat("TEMPORAL COVERAGE ANALYSIS:\n")
cat("Strategy A - Target Period Coverage:", round(temporal_A$target_period_coverage * 100, 1), "%\n")
cat("Strategy B - Target Period Coverage:", round(temporal_B$target_period_coverage * 100, 1), "%\n")

if (length(temporal_A$peak_years) > 0) {
  cat("Strategy A - Peak Publication Years:", paste(temporal_A$peak_years, collapse = ", "), "\n")
}
if (length(temporal_B$peak_years) > 0) {
  cat("Strategy B - Peak Publication Years:", paste(temporal_B$peak_years, collapse = ", "), "\n")
}

## ----performance_plots, fig.width=10, fig.height=6----------------------------
# Side-by-side performance overview
overview_A <- analyzer_A$visualize_performance("overview") +
  ggtitle("Strategy A: Broad COVID Research") +
  theme(plot.title = element_text(size = 12)) +
  ylim(0, 1)

overview_B <- analyzer_B$visualize_performance("overview") +
  ggtitle("Strategy B: Targeted Clinical COVID") +
  theme(plot.title = element_text(size = 12)) +
  ylim(0, 1)

if (requireNamespace("patchwork", quietly = TRUE)) {
  combined_overview <- overview_A + overview_B +
    plot_annotation(title = "COVID Search Strategy Performance Comparison",
                    subtitle = "Broad vs. Targeted Clinical Approaches")
  print(combined_overview)
} else {
  print(overview_A)
  print(overview_B)
}

## ----metric_comparison_plot, fig.width=8, fig.height=6------------------------
# Create direct comparison plot
comparison_data <- data.frame(
  Strategy = c("Broad COVID Research", "Targeted Clinical COVID"),
  Precision = c(metrics_A$precision_recall$precision, metrics_B$precision_recall$precision),
  Recall = c(metrics_A$precision_recall$recall, metrics_B$precision_recall$recall),
  F1_Score = c(metrics_A$precision_recall$f1_score, metrics_B$precision_recall$f1_score),
  stringsAsFactors = FALSE
)

# Reshape for plotting
comparison_long <- rbind(
  data.frame(Strategy = comparison_data$Strategy, Metric = "Precision", Value = comparison_data$Precision),
  data.frame(Strategy = comparison_data$Strategy, Metric = "Recall", Value = comparison_data$Recall),
  data.frame(Strategy = comparison_data$Strategy, Metric = "F1_Score", Value = comparison_data$F1_Score)
)

comparison_plot <- ggplot(comparison_long, aes(x = Metric, y = Value, fill = Strategy)) +
  geom_col(position = "dodge", alpha = 0.8, width = 0.7) +
  geom_text(aes(label = round(Value, 3)), position = position_dodge(width = 0.7), vjust = -0.5) +
  scale_fill_manual(values = c("Broad COVID Research" = "#2E86AB", "Targeted Clinical COVID" = "#A23B72")) +
  labs(title = "Direct Performance Metric Comparison",
       subtitle = "Trade-offs: Broad strategy shows higher recall, Targeted shows higher precision",
       y = "Score", x = "Performance Metric") +
  theme_minimal() +
  theme(legend.position = "bottom") +
  ylim(0, 1)

print(comparison_plot)

## ----overlap_plot, fig.width=8, fig.height=6----------------------------------
# Create overlap visualization
overlap_data <- data.frame(
  Category = c("Broad Strategy Only", "Overlap", "Targeted Strategy Only"),
  Count = c(strategy_comparison$overlap_analysis$unique_to_strategy1,
            strategy_comparison$overlap_analysis$overlap_count,
            strategy_comparison$overlap_analysis$unique_to_strategy2),
  stringsAsFactors = FALSE
)

overlap_data$Percentage <- overlap_data$Count / strategy_comparison$overlap_analysis$total_unique * 100

overlap_plot <- ggplot(overlap_data, aes(x = Category, y = Count, fill = Category)) +
  geom_col(alpha = 0.8, width = 0.7) +
  geom_text(aes(label = paste0(Count, "\n(", round(Percentage, 1), "%)")), vjust = 0.5, size = 3.5) +
  scale_fill_manual(values = c("Broad Strategy Only" = "#2E86AB",
                               "Overlap" = "#F18F01", 
                               "Targeted Strategy Only" = "#A23B72")) +
  labs(title = "Article Retrieval Overlap Analysis",
       subtitle = "Complementary nature of broad vs. targeted approaches",
       y = "Number of Articles", x = "Category") +
  theme_minimal() +
  theme(legend.position = "none",
        axis.text.x = element_text(angle = 45, hjust = 1))

print(overlap_plot)

## ----temporal_plots, fig.width=12, fig.height=6-------------------------------
# Temporal comparison
temporal_A_plot <- analyzer_A$visualize_performance("temporal") +
  ggtitle("Strategy A: Temporal Distribution") +
  theme(plot.title = element_text(size = 12))

temporal_B_plot <- analyzer_B$visualize_performance("temporal") +
  ggtitle("Strategy B: Temporal Distribution") +
  theme(plot.title = element_text(size = 12))

if (requireNamespace("patchwork", quietly = TRUE)) {
  combined_temporal <- temporal_A_plot + temporal_B_plot +
    plot_annotation(title = "Temporal Distribution Comparison")
  print(combined_temporal)
} else {
  print(temporal_A_plot)
  print(temporal_B_plot)
}

## ----term_effectiveness-------------------------------------------------------
# Analyze terms from Strategy A
term_analysis_A <- term_effectiveness(
  terms = strategy_A$terms,
  search_results = unique_A,
  gold_standard = gold_standard_ids,
  text_fields = c("title", "abstract")
)

# Analyze terms from Strategy B
term_analysis_B <- term_effectiveness(
  terms = strategy_B$terms,
  search_results = unique_B,
  gold_standard = gold_standard_ids,
  text_fields = c("title", "abstract")
)

# Calculate Term Effectiveness Scores (TES)
term_analysis_A <- calc_tes(term_analysis_A)
term_analysis_B <- calc_tes(term_analysis_B)

cat("Term Effectiveness for Strategy A (Broad COVID Research):\n")
print(term_analysis_A[, c("term", "precision", "coverage", "tes")])

cat("\nTerm Effectiveness for Strategy B (Targeted Clinical COVID):\n") 
print(term_analysis_B[, c("term", "precision", "coverage", "tes")])

## ----term_plots, fig.width=11, fig.height=8-----------------------------------
# Find top terms for both strategies
top_results_A <- find_top_terms(term_analysis_A, n = 3, plot = FALSE)
top_results_B <- find_top_terms(term_analysis_B, n = 3, plot = FALSE)

cat("Top performing terms in Strategy A:", paste(top_results_A$terms, collapse = ", "), "\n")
cat("Top performing terms in Strategy B:", paste(top_results_B$terms, collapse = ", "), "\n")

# Create precision plots for both strategies with better spacing
precision_plot_A <- plot_term_effectiveness(
  term_analysis_A,
  plot_type = "precision_only",
  title_override = "Strategy A: Term Precision Analysis",
  show_values = TRUE
) + theme(
  plot.margin = margin(20, 20, 20, 20),
  axis.text.y = element_text(size = 9),
  plot.title = element_text(size = 11)
)

precision_plot_B <- plot_term_effectiveness(
  term_analysis_B,
  plot_type = "precision_only", 
  title_override = "Strategy B: Term Precision Analysis",
  show_values = TRUE
) + theme(
  plot.margin = margin(20, 20, 20, 20),
  axis.text.y = element_text(size = 9),
  plot.title = element_text(size = 11)
)

if (requireNamespace("patchwork", quietly = TRUE)) {
  precision_comparison <- precision_plot_A + precision_plot_B +
    plot_annotation(
      title = "Term Precision Comparison Across Strategies",
      theme = theme(plot.title = element_text(size = 14, hjust = 0.5))
    )
  print(precision_comparison)
} else {
  print(precision_plot_A)
  print(precision_plot_B)
}

## ----bubble_plots, fig.width=12, fig.height=6---------------------------------
# Precision vs Coverage bubble plots
bubble_plot_A <- plot_term_effectiveness(
  term_analysis_A,
  plot_type = "precision_coverage",
  title_override = "Strategy A: Term Effectiveness Landscape",
  show_values = FALSE
)

bubble_plot_B <- plot_term_effectiveness(
  term_analysis_B,
  plot_type = "precision_coverage", 
  title_override = "Strategy B: Term Effectiveness Landscape",
  show_values = FALSE
)

if (requireNamespace("patchwork", quietly = TRUE)) {
  bubble_comparison <- bubble_plot_A + bubble_plot_B +
    plot_annotation(title = "Term Effectiveness Landscape Comparison")
  print(bubble_comparison)
} else {
  print(bubble_plot_A)
  print(bubble_plot_B)
}

## ----cross_strategy_terms-----------------------------------------------------
# Compare terms across strategies
term_comparison <- compare_terms(
  list(
    "Broad" = term_analysis_A,
    "Targeted" = term_analysis_B
  ),
  top_n = 3
)

cat("Cross-Strategy Term Effectiveness Comparison:\n")
print(term_comparison)

# Create TES comparison plot
top_terms_combined <- unique(c(top_results_A$terms, top_results_B$terms))
tes_comparison_data <- term_comparison[term_comparison$term %in% top_terms_combined, ]

tes_plot <- ggplot(tes_comparison_data, aes(x = term, y = tes, fill = strategy)) +
  geom_col(position = "dodge", alpha = 0.8, width = 0.7) +
  geom_text(aes(label = round(tes, 3)),
            position = position_dodge(width = 0.7), vjust = -0.5, size = 3) +
  scale_fill_manual(values = c("Broad" = "#2E86AB", "Targeted" = "#A23B72")) +
  labs(title = "Term Effectiveness Score (TES) Comparison",
       subtitle = "Top-performing terms across COVID search strategies",
       x = "Search Terms", y = "TES Score", fill = "Strategy") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1),
        legend.position = "bottom") +
  ylim(0, 1)

print(tes_plot)

## ----recommendations----------------------------------------------------------
# Determine winners by metric
precision_winner <- ifelse(metrics_A$precision_recall$precision > metrics_B$precision_recall$precision,
                           "Broad COVID Research (A)", "Targeted Clinical COVID (B)")
recall_winner <- ifelse(metrics_A$precision_recall$recall > metrics_B$precision_recall$recall,
                        "Broad COVID Research (A)", "Targeted Clinical COVID (B)")
f1_winner <- ifelse(metrics_A$precision_recall$f1_score > metrics_B$precision_recall$f1_score,
                    "Broad COVID Research (A)", "Targeted Clinical COVID (B)")

cat("PERFORMANCE WINNERS BY METRIC:\n")
cat("Best Precision:", precision_winner, "\n")
cat("Best Recall:", recall_winner, "\n") 
cat("Best Overall F1 Score:", f1_winner, "\n\n")

# Strategic recommendations
cat("USAGE RECOMMENDATIONS:\n")
cat("• For broad COVID-19 scoping reviews → Use Strategy A (Broad)\n")
cat("• For clinical intervention reviews → Use Strategy B (Targeted)\n")
cat("• For comprehensive systematic reviews → Combine both strategies\n")
cat("• For rapid evidence synthesis → Start with Strategy B, expand with Strategy A if needed\n\n")

# Complementarity assessment
if (strategy_comparison$overlap_analysis$overlap_percentage < 60) {
  cat("✓ HIGH COMPLEMENTARITY: Strategies are highly complementary - combining both is recommended\n")
} else {
  cat("○ MODERATE OVERLAP: Some redundancy between strategies\n")
}

if (strategy_comparison$complementarity$synergy_score > 0.15) {
  cat("✓ STRONG SYNERGY: Combining strategies provides substantial added value\n")
} else {
  cat("○ LIMITED SYNERGY: Minimal additional benefit from combining strategies\n")
}

## ----export_results-----------------------------------------------------------
# Create temporary directory for exports
output_dir <- tempdir()

# Export individual strategy results
export_files_A <- export_results(
  search_results = unique_A,
  file_path = file.path(output_dir, "strategy_A_broad_covid"),
  formats = c("csv", "xlsx"),
  include_metadata = TRUE
)

export_files_B <- export_results(
  search_results = unique_B,
  file_path = file.path(output_dir, "strategy_B_targeted_covid"),
  formats = c("csv", "xlsx"),
  include_metadata = TRUE
)

# Create comprehensive data package
enhanced_analysis_results <- list(
  metrics_A = metrics_A,
  metrics_B = metrics_B,
  comparison = comparison_result,
  strategy_comparison = strategy_comparison,
  temporal_A = temporal_A,
  temporal_B = temporal_B,
  term_effectiveness_A = term_analysis_A,
  term_effectiveness_B = term_analysis_B
)

package_dir <- create_data_package(
  search_results = rbind(
    transform(unique_A, search_strategy = "Broad_COVID_Research"),
    transform(unique_B, search_strategy = "Targeted_Clinical_COVID")
  ),
  analysis_results = enhanced_analysis_results,
  output_dir = output_dir,
  package_name = "covid_search_strategy_comparison"
)

cat("Analysis package created at:", package_dir, "\n")
cat("Individual exports created:\n")
for (file in c(export_files_A, export_files_B)) {
  cat("-", file, "\n")
}