Upload Phase1_analysis.py with huggingface_hub

Phase1_analysis.py

@@ -0,0 +1,103 @@
+import pandas as pd
+import numpy as np
+import pingouin as pg
+from semopy import Model
+from semopy import calc_stats
+
+# ---------------------------------------------------------
+# 1. LOAD DATA
+# ---------------------------------------------------------
+DATA_PATH = "phase1_clean.csv"
+print("Loading data from:", DATA_PATH)
+
+df = pd.read_csv(DATA_PATH)
+
+# ---------------------------------------------------------
+# 2. DEFINE OPERATORS
+# ---------------------------------------------------------
+operators = [
+    "accepting", "achieving", "appreciating", "arranging",
+    "boosting", "calculating", "constricting", "deciding",
+    "expanding", "sensing"
+]
+
+# ---------------------------------------------------------
+# 3. RECONSTRUCT ITEMS IN THE ORDER THEY APPEAR IN THE CSV
+# ---------------------------------------------------------
+ordered_items = {}
+
+for op in operators:
+    cols = [c for c in df.columns if c.startswith(op + "_")]
+    # Sort by the numeric suffix
+    cols_sorted = sorted(cols, key=lambda x: int(x.split("_")[1]))
+    ordered_items[op] = cols_sorted
+
+# ---------------------------------------------------------
+# 4. COMPUTE CRONBACH ALPHAS
+# ---------------------------------------------------------
+print("\n=== CRONBACH ALPHAS ===")
+alpha_results = {}
+
+for op, cols in ordered_items.items():
+    data = df[cols]
+    alpha = pg.cronbach_alpha(data)[0]
+    alpha_results[op] = alpha
+    print(f"{op:12s} α = {alpha:.3f}")
+
+pd.DataFrame(alpha_results.items(), columns=["operator", "alpha"]).to_csv(
+    "operator_reliability.csv", index=False
+)
+print("\nSaved operator_reliability.csv")
+
+# ---------------------------------------------------------
+# 5. BUILD CFA MODEL STRING
+# ---------------------------------------------------------
+model_lines = []
+for op, cols in ordered_items.items():
+    line = f"{op} =~ " + " + ".join(cols)
+    model_lines.append(line)
+
+model_string = "\n".join(model_lines)
+
+print("\n=== CFA MODEL STRING ===")
+print(model_string)
+
+# ---------------------------------------------------------
+# 6. FIT CFA MODEL
+# ---------------------------------------------------------
+print("\n=== FITTING CFA MODEL ===")
+
+# Drop rows with missing values
+all_items = [c for cols in ordered_items.values() for c in cols]
+cfa_data = df[all_items].dropna()
+
+model = Model(model_string)
+model.fit(cfa_data)
+
+# ---------------------------------------------------------
+# 7. FIT INDICES
+# ---------------------------------------------------------
+print("\n=== CFA FIT INDICES ===")
+stats = calc_stats(model)
+fit_df = pd.DataFrame(stats.items(), columns=["index", "value"])
+print(fit_df)
+
+fit_df.to_csv("cfa_fit_indices.csv", index=False)
+print("Saved cfa_fit_indices.csv")
+
+# ---------------------------------------------------------
+# 8. STANDARDIZED LOADINGS (modern semopy API)
+# ---------------------------------------------------------
+print("\n=== STANDARDIZED LOADINGS ===")
+
+params = model.inspect(std_est=True)
+
+# Filter for loadings: op == "=~"
+loadings = params[params["op"] == "~"]
+
+print(loadings)
+
+loadings.to_csv("cfa_loadings.csv", index=False)
+print("Saved cfa_loadings.csv")
+
+print("\nDone.")