chore: Update analysis modules and add docs gitignore

- Update CampaignAnalysis.jl, ExperimentCollector.jl, ResultsLoader.jl
- Add docs/Manifest.toml to .gitignore

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Author: gescholt

.gitignore

@@ -50,3 +50,4 @@ collected_experiments_*/
 
 # Experimental data (personal analysis, not package code)
 collected_experiments_*/
+docs/Manifest.toml

src/CampaignAnalysis.jl

@@ -181,8 +181,9 @@ function aggregate_csv_metrics_by_degree(campaign::CampaignResults)
     for exp in campaign.experiments
         exp_path = exp.source_path
 
-        # Find all CSV files for this experiment
-        csv_files = filter(f -> startswith(basename(f), "critical_points_deg_"),
+        # Find all CSV files for this experiment (both new and legacy formats)
+        csv_files = filter(f -> startswith(basename(f), "critical_points_raw_deg_") ||
+                               startswith(basename(f), "critical_points_deg_"),
                           readdir(exp_path, join=true, sort=false))
 
         for csv_file in csv_files
@@ -879,8 +880,9 @@ function print_parameter_recovery_table(campaign::CampaignResults)
             continue  # Skip if no true parameters
         end
 
-        # Load critical points by degree
-        csv_files = filter(f -> startswith(basename(f), "critical_points_deg_"),
+        # Load critical points by degree (both new and legacy formats)
+        csv_files = filter(f -> startswith(basename(f), "critical_points_raw_deg_") ||
+                               startswith(basename(f), "critical_points_deg_"),
                           readdir(exp_path, join=true))
 
         if isempty(csv_files)

src/ExperimentCollector.jl

@@ -155,7 +155,9 @@ function validate_experiment(exp_path::String)::ValidationResult
         files = readdir(exp_path)
 
         # Check for critical points CSV files
-        has_csv = any(f -> startswith(f, "critical_points_deg_") && endswith(f, ".csv"), files)
+        # Support both new format (critical_points_raw_deg_X.csv) and legacy (critical_points_deg_X.csv)
+        has_csv = any(f -> endswith(f, ".csv") &&
+            (startswith(f, "critical_points_raw_deg_") || startswith(f, "critical_points_deg_")), files)
 
         # Check for config file
         has_config = "experiment_config.json" in files

src/ResultsLoader.jl

@@ -51,7 +51,9 @@ function is_single_experiment(path::String)
     end
 
     files = readdir(path)
-    has_csv = any(f -> endswith(f, ".csv") && startswith(f, "critical_points_deg_"), files)
+    # Support both new format (critical_points_raw_deg_X.csv) and legacy (critical_points_deg_X.csv)
+    has_csv = any(f -> endswith(f, ".csv") &&
+        (startswith(f, "critical_points_raw_deg_") || startswith(f, "critical_points_deg_")), files)
     has_results = "results_summary.json" in files || "results_summary.jld2" in files
 
     return has_csv || has_results
@@ -389,7 +391,10 @@ function load_critical_points_from_csvs(dir_path::String, data::AbstractDict)
         degree_str = replace(string(degree_key), "degree_" => "")
         degree = parse(Int, degree_str)
 
-        csv_file = joinpath(dir_path, "critical_points_deg_$(degree).csv")
+        # Try new format first (Phase 2), fall back to legacy format
+        csv_file_raw = joinpath(dir_path, "critical_points_raw_deg_$(degree).csv")
+        csv_file_legacy = joinpath(dir_path, "critical_points_deg_$(degree).csv")
+        csv_file = isfile(csv_file_raw) ? csv_file_raw : csv_file_legacy
 
         if !isfile(csv_file)
             continue