Initial commit: AI code detection project (without binary files)

.gitignore

@@ -0,0 +1,30 @@
+# Binary model files
+*.pkl
+*.npy
+*.joblib
+
+# Cache files
+__pycache__/
+*.pyc
+*.pyo
+*.pyd
+.Python
+*.egg-info/
+dist/
+build/
+
+# IDE
+.vscode/
+.idea/
+*.swp
+*.swo
+
+# OS
+.DS_Store
+Thumbs.db
+
+# Data (optional - comment out if you want to include)
+dataset/raw/
+
+# State files  
+.states/

README.md

@@ -0,0 +1,43 @@
+---
+title: AI vs Human Code Classifier
+emoji: 🤖
+colorFrom: purple
+colorTo: blue
+sdk: gradio
+app_file: app.py
+python_version: "3.11.9"
+suggested_hardware: cpu-basic
+startup_duration_timeout: 60m
+preload_from_hub:
+  - microsoft/unixcoder-base
+models:
+  - microsoft/unixcoder-base
+short_description: Classify code as AI- or human-written (XGBoost + UnixCoder + SHAP).
+tags:
+  - code-classification
+  - explainability
+  - xgboost
+---
+
+# AI vs Human Code Classifier
+
+Gradio demo for [Hugging Face Spaces](https://huggingface.co/docs/hub/spaces-overview) using the same inference stack as the Reflex app (`reflex_ui`).
+
+## Deploy on Hugging Face
+
+1. Create a [new Space](https://huggingface.co/new-space), pick **Gradio**, then push this repository (or connect GitHub and set the repository).
+2. In **Settings → Hardware**, keep **CPU Basic** (free: 16 GB RAM) unless you need a GPU.
+3. Open the **App** tab. The first run may take a while while weights load.
+
+Public URL shape: `https://huggingface.co/spaces/<your-username>/<space-name>` (and the `.hf.space` embed URL from the Space menu).
+
+## Local Gradio (optional)
+
+```bash
+pip install -r requirements.txt
+python app.py
+```
+
+## Local Reflex UI
+
+See `reflex_ui/README.md` for the full Reflex front end.

app.py

@@ -0,0 +1,83 @@
+"""
+Hugging Face Spaces entrypoint (Gradio).
+Uses the same backend as the Reflex UI via `reflex_ui.backend_bridge.run_prediction`.
+"""
+
+from __future__ import annotations
+
+import os
+import sys
+from pathlib import Path
+
+import gradio as gr
+import pandas as pd
+
+REPO_ROOT = Path(__file__).resolve().parent
+os.chdir(REPO_ROOT)
+if str(REPO_ROOT) not in sys.path:
+    sys.path.insert(0, str(REPO_ROOT))
+
+_reflex_pkg_parent = REPO_ROOT / "reflex_ui"
+if str(_reflex_pkg_parent) not in sys.path:
+    sys.path.insert(0, str(_reflex_pkg_parent))
+
+from reflex_ui.backend_bridge import run_prediction  # noqa: E402
+
+
+def _predict(code: str, language: str, top_k: float):
+    text = (code or "").strip()
+    if not text:
+        return (
+            "*Paste a non-empty code snippet.*",
+            "",
+            pd.DataFrame(),
+            "",
+        )
+    lang = (language or "python").lower()
+    if lang not in ("python", "java"):
+        lang = "python"
+    k = int(top_k) if top_k else 6
+    k = max(1, min(30, k))
+    try:
+        r = run_prediction(code=text, language=lang, top_k=k)
+    except Exception as e:
+        return (f"**{type(e).__name__}:** `{e}`", "", pd.DataFrame(), "")
+
+    label_md = f"## Prediction: **{r.label}**\n\nP(AI) = **{r.prob_ai:.4f}**"
+    grouped_lines = [
+        f"- **{name}:** {val:.4f}" for name, val in sorted(r.grouped_importance.items(), key=lambda x: -x[1])
+    ]
+    grouped_md = "### Group importance\n" + ("\n".join(grouped_lines) if grouped_lines else "_n/a_")
+    df = pd.DataFrame(r.shap_rows)
+    if not df.empty and "impact" in df.columns:
+        show = df.drop(columns=["impact"], errors="ignore")
+    else:
+        show = df
+    return label_md, r.explanation, show, grouped_md
+
+
+with gr.Blocks(title="AI vs Human Code Classifier") as demo:
+    gr.Markdown(
+        "# AI vs Human Code Classifier\n"
+        "Classifies snippets as **AI** or **Human** using XGBoost on statistical, AST, "
+        "stylometry, language, and **UnixCoder** semantic features, with **SHAP** attributions."
+    )
+    with gr.Row():
+        language = gr.Radio(choices=["python", "java"], value="python", label="Language")
+        top_k = gr.Slider(1, 20, value=6, step=1, label="Top SHAP features")
+    code = gr.Code(label="Code", language="python", lines=18)
+    run_btn = gr.Button("Run prediction", variant="primary")
+    out_head = gr.Markdown()
+    out_grouped = gr.Markdown()
+    out_expl = gr.Markdown(label="Narrative explanation")
+    out_table = gr.Dataframe(label="Top feature attributions", wrap=True)
+
+    run_btn.click(
+        _predict,
+        inputs=[code, language, top_k],
+        outputs=[out_head, out_expl, out_table, out_grouped],
+    )
+
+
+if __name__ == "__main__":
+    demo.launch(server_name="0.0.0.0", server_port=7860)

basemodel/README.md

@@ -0,0 +1,3 @@
+This module implements a token‑level statistical baseline for AI vs human code detection. It relies solely on distributional properties of code tokens, such as entropy, burstiness, and predictability, without using structural or stylistic features. The goal is to evaluate whether LLM‑generated code exhibits detectable statistical signatures prior to incorporating higher‑level program structure and stylometry.
+
+The statistical baseline achieves 58.6% validation accuracy, indicating that token‑level regularities alone are insufficient for robust authorship discrimination, motivating the inclusion of structural and stylistic features.”

basemodel/evaluate_randomforest.py

@@ -0,0 +1,39 @@
+import pandas as pd
+import joblib
+from sklearn.metrics import accuracy_score, classification_report, confusion_matrix
+
+
+test_df = pd.read_csv("basemodel/test_features.csv")
+
+
+drop_cols = ["Label"]
+if "language" in test_df.columns:
+    drop_cols.append("language")
+
+X_test = test_df.drop(columns=drop_cols)
+y_test = test_df["Label"]
+
+
+rf = joblib.load("basemodel/random_forest_baseline.pkl")
+
+
+test_preds = rf.predict(X_test)
+
+accuracy = accuracy_score(y_test, test_preds)
+
+
+print("\n🧪 TEST SET EVALUATION (Statistical Baseline)\n")
+print("Test Accuracy:", round(accuracy, 4))
+
+print("\nClassification Report:\n")
+print(
+    classification_report(
+        y_test,
+        test_preds,
+        target_names=["Human", "AI"],
+        zero_division=0
+    )
+)
+
+print("\nConfusion Matrix:\n")
+print(confusion_matrix(y_test, test_preds))

basemodel/test_features.csv

@@ -0,0 +1,75 @@
+entropy,burstiness,repetition_ratio,avg_token_length,vocab_richness,num_tokens,unique_token_ratio,avg_line_length,language,Label
+4.83141993487849,0.781021897810219,0.583941605839416,3.153284671532847,0.41605839416058393,137,0.41605839416058393,30.09090909090909,python,1.0
+4.100105160750668,0.8333333333333334,0.6111111111111112,2.7962962962962963,0.3888888888888889,54,0.3888888888888889,33.5,python,0.0
+3.549523459597832,0.7857142857142857,0.5,6.5,0.5,28,0.5,27.818181818181817,java,1.0
+4.598877796136016,0.7543859649122807,0.49122807017543857,3.473684210526316,0.5087719298245614,57,0.5087719298245614,43.666666666666664,python,1.0
+4.008947073809811,0.7708333333333334,0.5625,2.5208333333333335,0.4375,48,0.4375,24.181818181818183,java,1.0
+3.9620439376607153,0.8448275862068966,0.6379310344827587,2.689655172413793,0.3620689655172414,58,0.3620689655172414,24.5,java,0.0
+4.496439344671016,0.65,0.375,4.325,0.625,40,0.625,41.0,python,1.0
+5.038474870244337,0.7261904761904762,0.5119047619047619,4.083333333333333,0.4880952380952381,84,0.4880952380952381,40.916666666666664,python,0.0
+3.828442584567261,0.9724770642201835,0.8256880733944953,2.8990825688073394,0.1743119266055046,109,0.1743119266055046,34.111111111111114,python,1.0
+5.158447453320608,0.7702702702702703,0.45945945945945943,4.581081081081081,0.5405405405405406,74,0.5405405405405406,19.52173913043478,java,0.0
+4.074700960493769,0.6470588235294118,0.5098039215686274,4.823529411764706,0.49019607843137253,51,0.49019607843137253,31.272727272727273,python,0.0
+5.216551170557516,0.32653061224489793,0.18367346938775508,6.489795918367347,0.8163265306122449,49,0.8163265306122449,31.466666666666665,java,1.0
+4.662553238462382,0.8235294117647058,0.5588235294117647,2.926470588235294,0.4411764705882353,68,0.4411764705882353,25.866666666666667,python,1.0
+4.718863069670588,0.8344827586206897,0.6827586206896552,4.703448275862069,0.31724137931034485,145,0.31724137931034485,28.0,java,0.0
+4.186133469144905,0.7878787878787878,0.6060606060606061,3.015151515151515,0.3939393939393939,66,0.3939393939393939,30.636363636363637,python,1.0
+4.201497374489525,0.9393939393939394,0.7474747474747474,1.9292929292929293,0.25252525252525254,99,0.25252525252525254,46.142857142857146,java,1.0
+5.555743206064364,0.7756410256410257,0.5769230769230769,4.397435897435898,0.4230769230769231,156,0.4230769230769231,33.875,java,0.0
+5.120767187212948,0.819672131147541,0.5819672131147541,4.122950819672131,0.4180327868852459,122,0.4180327868852459,27.517241379310345,java,0.0
+5.411695856960623,0.5733333333333334,0.3466666666666667,4.773333333333333,0.6533333333333333,75,0.6533333333333333,44.36363636363637,python,0.0
+5.9364666815595095,0.9323308270676691,0.7969924812030076,4.402255639097745,0.20300751879699247,532,0.20300751879699247,46.348314606741575,java,0.0
+5.921411473217016,0.8119266055045872,0.6146788990825688,5.041284403669724,0.3853211009174312,218,0.3853211009174312,43.567567567567565,python,0.0
+5.681975847739398,0.5217391304347826,0.3369565217391305,6.336956521739131,0.6630434782608695,92,0.6630434782608695,41.10526315789474,python,1.0
+4.9007043302212505,0.7849462365591398,0.5591397849462365,5.236559139784946,0.44086021505376344,93,0.44086021505376344,38.111111111111114,python,0.0
+5.3228285773092585,0.9466019417475728,0.7524271844660194,3.5194174757281553,0.24757281553398058,206,0.24757281553398058,33.48837209302326,java,0.0
+4.886224508551541,0.6811594202898551,0.4492753623188406,4.956521739130435,0.5507246376811594,69,0.5507246376811594,27.904761904761905,java,0.0
+3.7468422480983827,0.7317073170731707,0.5609756097560976,2.5853658536585367,0.43902439024390244,41,0.43902439024390244,23.1,python,1.0
+5.711349874129563,0.9204545454545454,0.7727272727272727,3.7017045454545454,0.22727272727272727,352,0.22727272727272727,33.15068493150685,java,0.0
+4.708644255945031,0.8160919540229885,0.6091954022988506,4.551724137931035,0.39080459770114945,87,0.39080459770114945,38.73684210526316,java,0.0
+4.366729296672175,0.46875,0.28125,6.21875,0.71875,32,0.71875,24.0,java,1.0
+4.819680984926536,0.7704918032786885,0.47540983606557374,3.19672131147541,0.5245901639344263,61,0.5245901639344263,33.18181818181818,python,0.0
+4.982067485400015,0.8495575221238938,0.6283185840707964,2.8849557522123894,0.37168141592920356,113,0.37168141592920356,37.5,python,0.0
+5.792349655949547,0.8646288209606987,0.6506550218340612,6.423580786026201,0.34934497816593885,229,0.34934497816593885,31.1,java,1.0
+5.172672355162728,0.8211382113821138,0.6016260162601625,3.2195121951219514,0.3983739837398374,123,0.3983739837398374,29.7,java,0.0
+4.509883902076781,0.5869565217391305,0.3913043478260869,3.9782608695652173,0.6086956521739131,46,0.6086956521739131,31.6,java,1.0
+5.0781577563714695,0.7582417582417582,0.5274725274725275,4.912087912087912,0.4725274725274725,91,0.4725274725274725,42.94736842105263,java,0.0
+4.752102792050065,0.8555555555555555,0.6222222222222222,2.6,0.37777777777777777,90,0.37777777777777777,30.875,python,1.0
+5.385167845762692,0.6842105263157895,0.4631578947368421,5.442105263157894,0.5368421052631579,95,0.5368421052631579,39.34782608695652,java,0.0
+4.698453748752097,0.8333333333333334,0.6111111111111112,2.3,0.3888888888888889,90,0.3888888888888889,30.75,python,1.0
+5.0936606896881855,0.48,0.26,4.5,0.74,50,0.74,36.44444444444444,python,1.0
+4.762058188070034,0.5689655172413793,0.3793103448275862,3.8793103448275863,0.6206896551724138,58,0.6206896551724138,35.875,python,0.0
+3.842216900031167,0.8703703703703703,0.6666666666666667,3.4444444444444446,0.3333333333333333,54,0.3333333333333333,26.571428571428573,python,1.0
+5.394236721618798,0.8340425531914893,0.6851063829787234,3.1659574468085108,0.3148936170212766,235,0.3148936170212766,45.16129032258065,java,0.0
+3.6368421881310113,0.5416666666666666,0.375,5.916666666666667,0.625,24,0.625,29.857142857142858,python,1.0
+4.371179616205941,0.821917808219178,0.6164383561643836,4.575342465753424,0.3835616438356164,73,0.3835616438356164,31.31578947368421,python,0.0
+5.292298716803247,0.6103896103896104,0.38961038961038963,4.597402597402597,0.6103896103896104,77,0.6103896103896104,43.63636363636363,python,0.0
+4.9019636334438,0.8170731707317073,0.5609756097560976,4.7926829268292686,0.43902439024390244,82,0.43902439024390244,34.21739130434783,java,0.0
+6.249267385727016,0.8602150537634409,0.6774193548387097,5.467741935483871,0.3225806451612903,372,0.3225806451612903,44.231884057971016,python,0.0
+5.022946255731872,0.905982905982906,0.6495726495726496,2.769230769230769,0.3504273504273504,117,0.3504273504273504,34.44,python,0.0
+5.183364648336087,0.578125,0.34375,6.140625,0.65625,64,0.65625,36.5,python,1.0
+3.8514475312208925,0.891566265060241,0.7469879518072289,2.5542168674698793,0.25301204819277107,83,0.25301204819277107,33.15384615384615,java,1.0
+4.76725052491479,0.8088235294117647,0.5294117647058824,4.382352941176471,0.47058823529411764,68,0.47058823529411764,29.857142857142858,java,0.0
+4.053149247727424,0.93,0.76,2.96,0.24,100,0.24,30.523809523809526,java,1.0
+5.508457113311519,0.8251748251748252,0.5944055944055944,5.1328671328671325,0.40559440559440557,143,0.40559440559440557,42.411764705882355,java,0.0
+4.041108871665959,0.7560975609756098,0.5121951219512195,5.048780487804878,0.4878048780487805,41,0.4878048780487805,26.75,java,1.0
+5.264706383768677,0.7769784172661871,0.5827338129496402,4.683453237410072,0.4172661870503597,139,0.4172661870503597,71.57142857142857,python,0.0
+4.888197350707755,0.9251700680272109,0.7210884353741497,2.5374149659863945,0.2789115646258503,147,0.2789115646258503,37.392857142857146,python,0.0
+4.246439344671016,0.65,0.42500000000000004,4.5,0.575,40,0.575,25.818181818181817,python,1.0
+5.3571067941182635,0.4838709677419355,0.27419354838709675,7.532258064516129,0.7258064516129032,62,0.7258064516129032,46.15384615384615,java,1.0
+4.882260151910624,0.8173076923076923,0.625,3.951923076923077,0.375,104,0.375,29.307692307692307,java,0.0
+4.413909765557392,0.40625,0.25,3.84375,0.75,32,0.75,28.666666666666668,python,0.0
+4.003044786734202,0.6666666666666666,0.4242424242424242,4.424242424242424,0.5757575757575758,33,0.5757575757575758,28.714285714285715,python,0.0
+4.138502179294739,0.8235294117647058,0.6176470588235294,3.8088235294117645,0.38235294117647056,68,0.38235294117647056,38.45454545454545,java,1.0
+3.490664945449586,0.8979591836734694,0.6938775510204082,4.571428571428571,0.30612244897959184,49,0.30612244897959184,26.526315789473685,java,0.0
+4.742155477497019,0.7049180327868853,0.47540983606557374,2.80327868852459,0.5245901639344263,61,0.5245901639344263,27.4,python,0.0
+4.907962016217061,0.4318181818181818,0.25,3.0454545454545454,0.75,44,0.75,62.0,python,1.0
+5.299012661837314,0.7024793388429752,0.5371900826446281,3.677685950413223,0.4628099173553719,121,0.4628099173553719,48.0,python,0.0
+5.643744827511846,0.8793774319066148,0.7081712062256809,7.603112840466926,0.2918287937743191,257,0.2918287937743191,36.523809523809526,java,1.0
+3.640223928941852,0.26666666666666666,0.1333333333333333,4.4,0.8666666666666667,15,0.8666666666666667,32.333333333333336,python,1.0
+4.220518125914522,0.9583333333333334,0.8194444444444444,2.0277777777777777,0.18055555555555555,144,0.18055555555555555,41.92307692307692,java,1.0
+5.706414205462762,0.7961783439490446,0.5796178343949044,7.503184713375796,0.42038216560509556,157,0.42038216560509556,38.61363636363637,java,1.0
+5.222228298430333,0.8230088495575221,0.5663716814159292,3.982300884955752,0.4336283185840708,113,0.4336283185840708,41.75,python,0.0
+4.277261292670126,0.8588235294117647,0.6705882352941177,3.0,0.32941176470588235,85,0.32941176470588235,34.35294117647059,python,0.0
+3.8035088547976783,0.5416666666666666,0.33333333333333337,4.208333333333333,0.6666666666666666,24,0.6666666666666666,24.571428571428573,python,1.0
+5.251158101143722,0.8026315789473685,0.618421052631579,4.894736842105263,0.3815789473684211,152,0.3815789473684211,44.06896551724138,java,0.0

basemodel/token_level_feature_extraction.py

@@ -0,0 +1,81 @@
+import pandas as pd
+import numpy as np
+import re
+import os
+from collections import Counter
+from math import log2
+
+
+def tokenize(code):
+    # identifiers, numbers, operators
+    return re.findall(r"[A-Za-z_]+|\d+|==|!=|<=|>=|[+\-*/%]", code)
+
+
+def token_entropy(tokens):
+    if not tokens:
+        return 0.0
+    counts = Counter(tokens)
+    total = len(tokens)
+    probs = [c / total for c in counts.values()]
+    return -sum(p * log2(p) for p in probs)
+
+def burstiness(tokens):
+    if not tokens:
+        return 0.0
+    counts = Counter(tokens)
+    repeated = sum(c for c in counts.values() if c > 1)
+    return repeated / len(tokens)
+
+def repetition_ratio(tokens):
+    if not tokens:
+        return 0.0
+    return 1 - (len(set(tokens)) / len(tokens))
+
+def avg_token_length(tokens):
+    if not tokens:
+        return 0.0
+    return np.mean([len(t) for t in tokens])
+
+def vocab_richness(tokens):
+    if not tokens:
+        return 0.0
+    return len(set(tokens)) / len(tokens)
+
+
+def extract_features(df):
+    features = []
+
+    for _, row in df.iterrows():
+        code = str(row["normalized_code"])
+        tokens = tokenize(code)
+
+        features.append({
+            "entropy": token_entropy(tokens),
+            "burstiness": burstiness(tokens),
+            "repetition_ratio": repetition_ratio(tokens),
+            "avg_token_length": avg_token_length(tokens),
+            "vocab_richness": vocab_richness(tokens),
+            "language": row.get("Language", "unknown")
+        })
+
+    return pd.DataFrame(features)
+
+
+if __name__ == "__main__":
+
+    os.makedirs("basemodel", exist_ok=True)
+
+    for split in ["train", "val", "test"]:
+        input_path = f"dataset/processed/dataset_{split}.csv"
+        df = pd.read_csv(input_path)
+
+        if "Label (0- HUMAN, 1-AI)" in df.columns:
+            df = df.rename(columns={"Label (0- HUMAN, 1-AI)": "Label"})
+
+        X = extract_features(df)
+        X["Label"] = df["Label"]
+
+        output_path = f"basemodel/{split}_features.csv"
+        X.to_csv(output_path, index=False)
+
+        print(f"Statistical baseline features extracted for {split}")

basemodel/train_features.csv

@@ -0,0 +1,345 @@
+entropy,burstiness,repetition_ratio,avg_token_length,vocab_richness,num_tokens,unique_token_ratio,avg_line_length,language,Label
+5.207956079154271,0.8712121212121212,0.6515151515151515,3.9015151515151514,0.3484848484848485,132,0.3484848484848485,41.464285714285715,python,0.0
+4.844227218082805,0.8962962962962963,0.7037037037037037,3.1259259259259258,0.2962962962962963,135,0.2962962962962963,15.704545454545455,java,0.0
+5.110041502517433,0.916083916083916,0.6783216783216783,4.5174825174825175,0.32167832167832167,143,0.32167832167832167,25.784313725490197,java,0.0
+4.599986090068817,0.8839285714285714,0.7053571428571428,4.616071428571429,0.29464285714285715,112,0.29464285714285715,30.68421052631579,java,0.0
+4.469392725378344,0.8589743589743589,0.6153846153846154,3.230769230769231,0.38461538461538464,78,0.38461538461538464,23.923076923076923,java,0.0
+4.293744383186907,0.859375,0.625,2.953125,0.375,64,0.375,26.210526315789473,java,0.0
+4.549676205034737,0.860655737704918,0.6885245901639344,5.918032786885246,0.3114754098360656,122,0.3114754098360656,33.96666666666667,python,0.0
+4.328596578632476,0.8163265306122449,0.5306122448979591,3.2244897959183674,0.46938775510204084,49,0.46938775510204084,28.333333333333332,python,0.0
+4.597753650865813,0.7941176470588235,0.5588235294117647,3.0441176470588234,0.4411764705882353,68,0.4411764705882353,30.916666666666668,java,1.0
+4.0675361498812554,0.8333333333333334,0.6666666666666667,2.0384615384615383,0.3333333333333333,78,0.3333333333333333,28.0,python,0.0
+4.853956434293166,0.898989898989899,0.6464646464646464,4.101010101010101,0.35353535353535354,99,0.35353535353535354,31.40740740740741,java,0.0
+6.10213285294983,0.875,0.6736111111111112,7.423611111111111,0.3263888888888889,288,0.3263888888888889,37.31395348837209,java,1.0
+3.053791405126089,0.7777777777777778,0.6388888888888888,4.027777777777778,0.3611111111111111,36,0.3611111111111111,22.25,python,1.0
+4.379556754405944,0.8620689655172413,0.5689655172413793,4.120689655172414,0.43103448275862066,58,0.43103448275862066,35.36363636363637,java,1.0
+4.538844415918181,0.8026315789473685,0.5921052631578947,3.3026315789473686,0.40789473684210525,76,0.40789473684210525,15.043478260869565,java,0.0
+4.435455286899907,0.847457627118644,0.576271186440678,2.694915254237288,0.423728813559322,59,0.423728813559322,26.818181818181817,python,1.0
+4.1277054819279435,0.8333333333333334,0.5625,3.0,0.4375,48,0.4375,24.6,java,1.0
+4.498419401169,0.7058823529411765,0.47058823529411764,3.2941176470588234,0.5294117647058824,51,0.5294117647058824,25.866666666666667,java,0.0
+3.615922063835167,0.7333333333333333,0.5,5.433333333333334,0.5,30,0.5,30.125,python,0.0
+5.043983303761563,0.5833333333333334,0.3666666666666667,4.983333333333333,0.6333333333333333,60,0.6333333333333333,37.5,python,1.0
+3.6897037321995474,0.42857142857142855,0.2857142857142857,7.095238095238095,0.7142857142857143,21,0.7142857142857143,22.7,java,1.0
+3.791950834682379,0.7142857142857143,0.5934065934065934,2.769230769230769,0.4065934065934066,91,0.4065934065934066,34.0,python,1.0
+5.542412227395111,0.7934782608695652,0.6141304347826086,3.6793478260869565,0.3858695652173913,184,0.3858695652173913,61.75,java,0.0
+3.8463189626846366,0.8947368421052632,0.5526315789473684,3.0526315789473686,0.4473684210526316,38,0.4473684210526316,20.6,python,1.0
+4.7720552088742005,0.6666666666666666,0.375,6.229166666666667,0.625,48,0.625,29.214285714285715,java,1.0
+5.085703343051975,0.7099236641221374,0.549618320610687,3.931297709923664,0.45038167938931295,131,0.45038167938931295,32.40909090909091,python,0.0
+4.253863368422077,0.6578947368421053,0.42105263157894735,3.6052631578947367,0.5789473684210527,38,0.5789473684210527,34.833333333333336,java,1.0
+5.285942953673669,0.8157894736842105,0.5614035087719298,3.4035087719298245,0.43859649122807015,114,0.43859649122807015,45.666666666666664,java,0.0
+4.324534762707879,0.7111111111111111,0.4666666666666667,8.177777777777777,0.5333333333333333,45,0.5333333333333333,37.30769230769231,java,1.0
+4.895714600514266,0.8457943925233645,0.7383177570093458,2.6448598130841123,0.2616822429906542,214,0.2616822429906542,18.784313725490197,java,0.0
+3.327339900083509,0.6666666666666666,0.4285714285714286,3.0952380952380953,0.5714285714285714,21,0.5714285714285714,22.8,python,1.0
+5.581633377149314,0.5934065934065934,0.37362637362637363,6.318681318681318,0.6263736263736264,91,0.6263736263736264,40.78947368421053,python,1.0
+5.657995400316549,0.8997289972899729,0.7425474254742548,3.907859078590786,0.25745257452574527,369,0.25745257452574527,56.26315789473684,java,0.0
+4.029187712039474,0.8367346938775511,0.5918367346938775,2.9183673469387754,0.40816326530612246,49,0.40816326530612246,30.0,java,1.0
+4.8582227682726,0.9393939393939394,0.7348484848484849,2.6363636363636362,0.26515151515151514,132,0.26515151515151514,32.0,python,1.0
+4.42309119581308,0.7941176470588235,0.5735294117647058,2.5588235294117645,0.4264705882352941,68,0.4264705882352941,30.454545454545453,java,1.0
+3.7665853954588724,0.7704918032786885,0.639344262295082,2.8852459016393444,0.36065573770491804,61,0.36065573770491804,33.90909090909091,python,0.0
+5.197908393371423,0.8072289156626506,0.6325301204819277,5.0843373493975905,0.3674698795180723,166,0.3674698795180723,38.65625,java,0.0
+4.536286231168867,0.42857142857142855,0.2571428571428571,3.6,0.7428571428571429,35,0.7428571428571429,31.166666666666668,python,1.0
+4.709586788340617,0.8957055214723927,0.7300613496932515,3.361963190184049,0.26993865030674846,163,0.26993865030674846,17.816326530612244,java,0.0
+4.5049847535671015,0.9545454545454546,0.7897727272727273,3.3806818181818183,0.21022727272727273,176,0.21022727272727273,36.16129032258065,java,0.0
+4.9423908406293835,0.6538461538461539,0.4358974358974359,3.8333333333333335,0.5641025641025641,78,0.5641025641025641,45.125,python,0.0
+4.41120441892766,0.7777777777777778,0.4666666666666667,6.488888888888889,0.5333333333333333,45,0.5333333333333333,44.77777777777778,python,0.0
+5.262208612706845,0.7327586206896551,0.5431034482758621,4.146551724137931,0.45689655172413796,116,0.45689655172413796,28.666666666666668,python,1.0
+4.208410187268525,0.48148148148148145,0.2592592592592593,4.222222222222222,0.7407407407407407,27,0.7407407407407407,32.2,python,1.0
+5.793376775298043,0.8823529411764706,0.680672268907563,5.529411764705882,0.31932773109243695,238,0.31932773109243695,27.633333333333333,java,1.0
+5.691854216621956,0.7849462365591398,0.5967741935483871,4.973118279569892,0.4032258064516129,186,0.4032258064516129,40.333333333333336,python,0.0
+5.058984089445427,0.3333333333333333,0.16666666666666663,6.785714285714286,0.8333333333333334,42,0.8333333333333334,30.923076923076923,python,1.0
+4.433157434964864,0.9736842105263158,0.7543859649122807,2.4035087719298245,0.24561403508771928,114,0.24561403508771928,31.5,python,1.0
+5.268123017576671,0.6164383561643836,0.3835616438356164,4.917808219178082,0.6164383561643836,73,0.6164383561643836,44.09090909090909,python,0.0
+6.198108865484992,0.7208121827411168,0.5025380710659899,5.0456852791878175,0.49746192893401014,197,0.49746192893401014,41.638888888888886,python,1.0
+4.5615159206608356,0.8227848101265823,0.6075949367088608,3.5569620253164556,0.3924050632911392,79,0.3924050632911392,13.678571428571429,java,0.0
+5.04371468310881,0.6351351351351351,0.43243243243243246,4.1891891891891895,0.5675675675675675,74,0.5675675675675675,37.0,java,0.0
+3.995906598484245,0.6944444444444444,0.4722222222222222,4.333333333333333,0.5277777777777778,36,0.5277777777777778,27.5,python,1.0
+4.757950193468896,0.9148936170212766,0.648936170212766,2.893617021276596,0.35106382978723405,94,0.35106382978723405,34.0,python,1.0
+5.053560274699695,0.9136690647482014,0.7050359712230216,4.705035971223022,0.2949640287769784,139,0.2949640287769784,39.69565217391305,python,0.0
+3.833323296470577,0.8813559322033898,0.6779661016949152,2.288135593220339,0.3220338983050847,59,0.3220338983050847,27.5,java,1.0
+4.253212018409155,0.5882352941176471,0.3529411764705882,4.117647058823529,0.6470588235294118,34,0.6470588235294118,13.75,java,0.0
+4.611949334080443,0.5384615384615384,0.3076923076923077,4.717948717948718,0.6923076923076923,39,0.6923076923076923,34.42857142857143,python,0.0
+4.756182308560408,0.9186991869918699,0.7073170731707317,3.3089430894308944,0.2926829268292683,123,0.2926829268292683,28.09090909090909,java,0.0
+5.1256959966316895,0.9022988505747126,0.6954022988505747,3.8045977011494254,0.3045977011494253,174,0.3045977011494253,31.727272727272727,java,0.0
+4.274386660093705,0.8103448275862069,0.5862068965517242,2.2413793103448274,0.41379310344827586,58,0.41379310344827586,39.166666666666664,java,1.0
+5.649232493101554,0.8258064516129032,0.5870967741935484,5.961290322580645,0.4129032258064516,155,0.4129032258064516,29.104166666666668,java,0.0
+3.8005595576361078,0.875,0.6785714285714286,3.0892857142857144,0.32142857142857145,56,0.32142857142857145,24.357142857142858,python,1.0
+4.635290969626768,0.6666666666666666,0.43137254901960786,3.7254901960784315,0.5686274509803921,51,0.5686274509803921,25.46153846153846,java,0.0
+4.854346937661789,0.5740740740740741,0.37037037037037035,5.0,0.6296296296296297,54,0.6296296296296297,25.4,java,0.0
+4.393061650825727,0.5,0.2941176470588235,7.352941176470588,0.7058823529411765,34,0.7058823529411765,40.625,python,1.0
+4.422442136473173,0.5135135135135135,0.32432432432432434,5.081081081081081,0.6756756756756757,37,0.6756756756756757,39.166666666666664,python,1.0
+4.207435516759526,0.6571428571428571,0.4,3.0285714285714285,0.6,35,0.6,22.375,python,1.0
+5.99203259332159,0.7687074829931972,0.48299319727891155,5.122448979591836,0.5170068027210885,147,0.5170068027210885,39.357142857142854,python,1.0
+6.007074413338004,0.711764705882353,0.5058823529411764,4.129411764705883,0.49411764705882355,170,0.49411764705882355,58.0,python,0.0
+3.6464393446710157,0.5,0.30000000000000004,3.85,0.7,20,0.7,27.0,python,1.0
+4.497874075147441,0.5333333333333333,0.37777777777777777,4.311111111111111,0.6222222222222222,45,0.6222222222222222,30.636363636363637,java,1.0
+4.728255941527934,0.9318181818181818,0.7348484848484849,2.6363636363636362,0.26515151515151514,132,0.26515151515151514,33.68,python,0.0
+4.2354799985372935,0.6862745098039216,0.5098039215686274,2.9411764705882355,0.49019607843137253,51,0.49019607843137253,28.333333333333332,python,0.0
+5.560363930440956,0.8622754491017964,0.6347305389221557,4.029940119760479,0.3652694610778443,167,0.3652694610778443,31.70212765957447,java,0.0
+6.014883786593109,0.5323741007194245,0.38129496402877694,4.863309352517986,0.6187050359712231,139,0.6187050359712231,34.25,python,1.0
+4.384183719779189,0.55,0.375,6.35,0.625,40,0.625,32.916666666666664,java,1.0
+4.36346329043209,0.7105263157894737,0.39473684210526316,6.157894736842105,0.6052631578947368,38,0.6052631578947368,27.733333333333334,java,1.0
+5.92272462852426,0.7397959183673469,0.5510204081632653,4.969387755102041,0.4489795918367347,196,0.4489795918367347,72.89473684210526,java,0.0
+4.17415473143563,0.9078947368421053,0.7105263157894737,2.1578947368421053,0.2894736842105263,76,0.2894736842105263,30.692307692307693,python,1.0
+5.271209951802516,0.9140271493212669,0.7330316742081449,4.97737556561086,0.2669683257918552,221,0.2669683257918552,43.416666666666664,java,0.0
+5.014527492014123,0.9281045751633987,0.7320261437908497,2.7777777777777777,0.2679738562091503,153,0.2679738562091503,38.869565217391305,python,0.0
+4.482483990770269,0.9565217391304348,0.7391304347826086,4.026086956521739,0.2608695652173913,115,0.2608695652173913,36.90909090909091,java,1.0
+5.722786643861381,0.8493975903614458,0.5783132530120482,3.7710843373493974,0.42168674698795183,166,0.42168674698795183,19.595744680851062,java,0.0
+5.300160551498397,0.5909090909090909,0.33333333333333337,5.757575757575758,0.6666666666666666,66,0.6666666666666666,36.75,java,1.0
+4.45148902643068,0.5,0.3157894736842105,3.4210526315789473,0.6842105263157895,38,0.6842105263157895,21.7,python,1.0
+5.072571258365157,0.8296703296703297,0.6428571428571428,3.4450549450549453,0.35714285714285715,182,0.35714285714285715,28.515151515151516,python,1.0
+4.168243795012672,0.5161290322580645,0.32258064516129037,6.0,0.6774193548387096,31,0.6774193548387096,31.375,java,0.0
+4.123939173306749,0.6511627906976745,0.4651162790697675,3.6744186046511627,0.5348837209302325,43,0.5348837209302325,28.2,java,1.0
+4.534110630792036,0.8888888888888888,0.6666666666666667,3.1222222222222222,0.3333333333333333,90,0.3333333333333333,35.3,java,0.0
+4.889260316971673,1.0,0.7348484848484849,3.6515151515151514,0.26515151515151514,132,0.26515151515151514,23.066666666666666,java,0.0
+4.883002384777129,0.9366197183098591,0.7323943661971831,2.852112676056338,0.2676056338028169,142,0.2676056338028169,36.25806451612903,python,0.0
+3.8200705941688406,0.9666666666666667,0.8111111111111111,5.011111111111111,0.18888888888888888,90,0.18888888888888888,37.421052631578945,java,1.0
+5.866907260896749,0.7725321888412017,0.5965665236051503,5.141630901287554,0.4034334763948498,233,0.4034334763948498,71.04347826086956,python,0.0
+4.592226257370053,0.7936507936507936,0.5396825396825398,3.5714285714285716,0.4603174603174603,63,0.4603174603174603,31.5,python,0.0
+4.979264809390598,0.2926829268292683,0.1707317073170732,6.317073170731708,0.8292682926829268,41,0.8292682926829268,33.416666666666664,java,1.0
+3.1693150678026982,0.8705882352941177,0.7294117647058824,2.1411764705882352,0.27058823529411763,85,0.27058823529411763,29.88888888888889,python,1.0
+4.718291773866265,0.8803418803418803,0.6837606837606838,2.871794871794872,0.3162393162393162,117,0.3162393162393162,33.17857142857143,python,0.0
+6.057756669289524,0.5080645161290323,0.3306451612903226,5.588709677419355,0.6693548387096774,124,0.6693548387096774,37.57692307692308,python,1.0
+4.94407911950056,0.8181818181818182,0.5324675324675325,7.077922077922078,0.4675324675324675,77,0.4675324675324675,37.05,java,1.0
+4.726143834398879,0.6136363636363636,0.34090909090909094,4.840909090909091,0.6590909090909091,44,0.6590909090909091,28.25,python,1.0
+4.475969855129791,0.8305084745762712,0.5593220338983051,3.389830508474576,0.4406779661016949,59,0.4406779661016949,27.0,java,1.0
+5.689839588541731,0.783410138248848,0.6082949308755761,3.640552995391705,0.391705069124424,217,0.391705069124424,31.355555555555554,java,0.0
+4.518215514675149,0.9473684210526315,0.6578947368421053,2.526315789473684,0.34210526315789475,76,0.34210526315789475,44.285714285714285,python,1.0
+4.124232718690979,0.7678571428571429,0.5892857142857143,3.1607142857142856,0.4107142857142857,56,0.4107142857142857,25.3125,java,0.0
+4.112426730000285,0.8846153846153846,0.7403846153846154,2.951923076923077,0.25961538461538464,104,0.25961538461538464,34.09090909090909,python,0.0
+5.039480564148708,0.7866666666666666,0.4933333333333333,4.093333333333334,0.5066666666666667,75,0.5066666666666667,32.111111111111114,java,0.0
+4.677139104850461,0.8181818181818182,0.5454545454545454,3.621212121212121,0.45454545454545453,66,0.45454545454545453,22.0,python,0.0
+4.131300342505361,0.7096774193548387,0.3870967741935484,6.161290322580645,0.6129032258064516,31,0.6129032258064516,31.625,java,1.0
+5.462480227696889,0.7959183673469388,0.5850340136054422,4.707482993197279,0.41496598639455784,147,0.41496598639455784,24.872340425531913,java,0.0
+3.6914280318460246,0.8125,0.53125,2.09375,0.46875,32,0.46875,23.333333333333332,python,1.0
+4.214114119970501,0.7142857142857143,0.5102040816326531,2.816326530612245,0.4897959183673469,49,0.4897959183673469,22.583333333333332,java,1.0
+5.697217023615801,0.8171428571428572,0.5828571428571429,5.154285714285714,0.41714285714285715,175,0.41714285714285715,37.93023255813954,java,0.0
+4.0791433740260095,0.6571428571428571,0.4285714285714286,5.057142857142857,0.5714285714285714,35,0.5714285714285714,25.0,python,1.0
+3.7489948035250964,0.47619047619047616,0.2857142857142857,3.4285714285714284,0.7142857142857143,21,0.7142857142857143,26.25,python,1.0
+4.907735246015272,0.75,0.525,4.2125,0.475,80,0.475,29.869565217391305,python,0.0
+4.822037278990095,0.7808219178082192,0.5342465753424658,2.958904109589041,0.4657534246575342,73,0.4657534246575342,33.0,python,0.0
+5.030104271105216,0.8301886792452831,0.6037735849056604,3.349056603773585,0.39622641509433965,106,0.39622641509433965,33.958333333333336,java,1.0
+4.816108939837481,0.7142857142857143,0.4285714285714286,4.392857142857143,0.5714285714285714,56,0.5714285714285714,29.23076923076923,python,1.0
+2.4276042408847402,0.8461538461538461,0.7230769230769231,1.7692307692307692,0.27692307692307694,65,0.27692307692307694,27.666666666666668,python,1.0
+6.584083875285213,0.8148148148148148,0.6271604938271604,6.212345679012346,0.3728395061728395,405,0.3728395061728395,40.2247191011236,python,0.0
+4.1673958618292355,0.8536585365853658,0.6951219512195121,2.7560975609756095,0.3048780487804878,82,0.3048780487804878,29.65,java,0.0
+3.741962789037907,0.6896551724137931,0.4482758620689655,2.0689655172413794,0.5517241379310345,29,0.5517241379310345,18.285714285714285,python,1.0
+4.86754794914527,0.5178571428571429,0.3571428571428571,3.375,0.6428571428571429,56,0.6428571428571429,19.2,java,0.0
+3.9331695057515157,0.9285714285714286,0.5952380952380952,3.5714285714285716,0.40476190476190477,42,0.40476190476190477,38.166666666666664,python,1.0
+4.196483601699262,0.49019607843137253,0.4117647058823529,6.254901960784314,0.5882352941176471,51,0.5882352941176471,34.25,python,0.0
+5.906626836308724,0.8967254408060453,0.7481108312342569,3.44080604534005,0.2518891687657431,397,0.2518891687657431,53.51219512195122,java,0.0
+6.369927477083823,0.5966850828729282,0.4088397790055248,4.900552486187845,0.5911602209944752,181,0.5911602209944752,49.65384615384615,python,1.0
+4.384795487373134,0.5277777777777778,0.33333333333333337,4.611111111111111,0.6666666666666666,36,0.6666666666666666,18.25,java,0.0
+5.139173834043169,0.8348623853211009,0.5871559633027523,4.935779816513762,0.41284403669724773,109,0.41284403669724773,35.370370370370374,java,0.0
+4.238521032323433,0.7708333333333334,0.5208333333333333,4.5625,0.4791666666666667,48,0.4791666666666667,27.571428571428573,python,0.0
+4.460619643646408,0.8309859154929577,0.5915492957746479,4.704225352112676,0.4084507042253521,71,0.4084507042253521,35.72222222222222,java,0.0
+5.2878758225325395,0.9408450704225352,0.8169014084507042,4.366197183098592,0.18309859154929578,355,0.18309859154929578,42.14545454545455,java,0.0
+4.8335434078297626,0.6666666666666666,0.4545454545454546,4.46969696969697,0.5454545454545454,66,0.5454545454545454,65.66666666666667,python,0.0
+3.3889821967054377,0.9555555555555556,0.7333333333333334,5.022222222222222,0.26666666666666666,45,0.26666666666666666,53.0,java,1.0
+5.626267512421823,0.8212290502793296,0.6145251396648045,3.8491620111731844,0.3854748603351955,179,0.3854748603351955,55.166666666666664,python,1.0
+4.5054500825687525,0.9090909090909091,0.696969696969697,3.6464646464646466,0.30303030303030304,99,0.30303030303030304,41.36363636363637,python,0.0
+4.66816999226466,0.5609756097560976,0.31707317073170727,4.7560975609756095,0.6829268292682927,41,0.6829268292682927,30.2,java,1.0
+4.1138276215159735,0.8461538461538461,0.6307692307692307,2.876923076923077,0.36923076923076925,65,0.36923076923076925,31.071428571428573,java,0.0
+3.7489948035250964,0.47619047619047616,0.2857142857142857,6.904761904761905,0.7142857142857143,21,0.7142857142857143,33.57142857142857,java,1.0
+4.639932288803049,0.8472222222222222,0.6944444444444444,3.9166666666666665,0.3055555555555556,144,0.3055555555555556,40.34615384615385,java,0.0
+5.503071993063162,0.7610619469026548,0.504424778761062,3.2920353982300883,0.49557522123893805,113,0.49557522123893805,51.36363636363637,python,1.0
+5.320245758154888,0.8130081300813008,0.5853658536585367,2.227642276422764,0.4146341463414634,123,0.4146341463414634,33.5625,python,1.0
+4.448860987743577,0.5789473684210527,0.3421052631578947,4.131578947368421,0.6578947368421053,38,0.6578947368421053,32.125,python,1.0
+4.526180668235663,0.5897435897435898,0.33333333333333337,4.17948717948718,0.6666666666666666,39,0.6666666666666666,26.9,java,1.0
+4.202589101773195,0.9411764705882353,0.7294117647058824,1.7058823529411764,0.27058823529411763,85,0.27058823529411763,31.3,java,1.0
+5.934435083287192,0.6229508196721312,0.4016393442622951,5.0,0.5983606557377049,122,0.5983606557377049,36.0,python,1.0
+4.713707321865714,0.6885245901639344,0.4590163934426229,6.459016393442623,0.5409836065573771,61,0.5409836065573771,33.8125,python,0.0
+4.3180873032876175,0.9,0.7111111111111111,2.2777777777777777,0.28888888888888886,90,0.28888888888888886,38.833333333333336,python,0.0
+4.285367908896077,0.6875,0.47916666666666663,4.520833333333333,0.5208333333333334,48,0.5208333333333334,40.875,java,1.0
+4.366596139176847,0.6585365853658537,0.41463414634146345,5.609756097560975,0.5853658536585366,41,0.5853658536585366,34.8,java,1.0
+5.343782087298768,0.8652482269503546,0.624113475177305,5.7304964539007095,0.375886524822695,141,0.375886524822695,40.40625,java,0.0
+4.327819531114783,0.375,0.25,6.28125,0.75,32,0.75,29.5,java,1.0
+3.8033524347703986,0.7105263157894737,0.5,2.763157894736842,0.5,38,0.5,26.375,python,0.0
+5.72576174609731,0.797979797979798,0.6161616161616161,3.984848484848485,0.3838383838383838,198,0.3838383838383838,42.55882352941177,java,0.0
+5.409509758627718,0.841025641025641,0.676923076923077,5.671794871794872,0.3230769230769231,195,0.3230769230769231,34.90384615384615,python,0.0
+5.679736241553765,0.7013888888888888,0.5138888888888888,4.743055555555555,0.4861111111111111,144,0.4861111111111111,52.55555555555556,python,0.0
+4.4912054473944405,0.8148148148148148,0.5185185185185186,5.277777777777778,0.48148148148148145,54,0.48148148148148145,36.23076923076923,java,0.0
+3.9362434129830604,0.6333333333333333,0.4,3.1333333333333333,0.6,30,0.6,24.75,python,0.0
+4.215928174708332,0.8852459016393442,0.6229508196721312,2.721311475409836,0.3770491803278688,61,0.3770491803278688,27.923076923076923,python,0.0
+5.255072207410051,0.9081081081081082,0.7135135135135136,3.675675675675676,0.2864864864864865,185,0.2864864864864865,38.34375,python,0.0
+4.8132935840925875,0.8902439024390244,0.5853658536585367,4.365853658536586,0.4146341463414634,82,0.4146341463414634,39.166666666666664,python,0.0
+3.728999521334141,0.9047619047619048,0.7301587301587302,4.619047619047619,0.2698412698412698,63,0.2698412698412698,34.84615384615385,java,1.0
+4.4003105301349255,0.6888888888888889,0.4444444444444444,2.5555555555555554,0.5555555555555556,45,0.5555555555555556,21.90909090909091,python,1.0
+5.192613181399552,0.8343949044585988,0.6624203821656052,3.445859872611465,0.3375796178343949,157,0.3375796178343949,42.476190476190474,python,0.0
+4.896037310447887,0.8761061946902655,0.6548672566371682,3.2123893805309733,0.34513274336283184,113,0.34513274336283184,27.653846153846153,java,0.0
+4.652782517460112,0.6447368421052632,0.513157894736842,2.8684210526315788,0.4868421052631579,76,0.4868421052631579,32.76470588235294,java,0.0
+3.757539644379974,0.9032258064516129,0.7096774193548387,1.6612903225806452,0.2903225806451613,62,0.2903225806451613,24.375,java,1.0
+5.4053549514703,0.7605633802816901,0.5633802816901409,5.640845070422535,0.43661971830985913,142,0.43661971830985913,38.833333333333336,java,0.0
+4.715163766425403,0.9142857142857143,0.6761904761904762,4.580952380952381,0.3238095238095238,105,0.3238095238095238,27.733333333333334,java,0.0
+5.048749755702801,0.8085106382978723,0.5638297872340425,2.74468085106383,0.43617021276595747,94,0.43617021276595747,34.375,java,1.0
+4.412142047769037,0.734375,0.546875,2.9375,0.453125,64,0.453125,31.0,python,0.0
+4.162294909570876,0.4444444444444444,0.2592592592592593,4.148148148148148,0.7407407407407407,27,0.7407407407407407,21.428571428571427,java,0.0
+5.17784699780413,0.46153846153846156,0.32307692307692304,3.4153846153846152,0.676923076923077,65,0.676923076923077,38.92307692307692,java,0.0
+5.4865950410391395,0.8578199052132701,0.6872037914691943,4.1753554502369665,0.3127962085308057,211,0.3127962085308057,37.65217391304348,java,0.0
+5.104500026854035,0.875,0.6333333333333333,4.291666666666667,0.36666666666666664,120,0.36666666666666664,39.80769230769231,java,0.0
+6.146771906965858,0.8115183246073299,0.5445026178010471,5.350785340314136,0.45549738219895286,191,0.45549738219895286,40.8,python,0.0
+4.898714188796334,0.8387096774193549,0.6021505376344086,4.559139784946237,0.3978494623655914,93,0.3978494623655914,40.0,java,0.0
+4.33379987011787,0.6521739130434783,0.4565217391304348,4.195652173913044,0.5434782608695652,46,0.5434782608695652,21.944444444444443,java,0.0
+4.720077865655612,0.8817204301075269,0.6344086021505376,4.806451612903226,0.3655913978494624,93,0.3655913978494624,35.95238095238095,java,0.0
+4.971145882958361,0.543859649122807,0.3508771929824561,4.0701754385964914,0.6491228070175439,57,0.6491228070175439,40.375,python,1.0
+4.608465630737972,0.89,0.6799999999999999,2.32,0.32,100,0.32,30.428571428571427,java,1.0
+4.553348286183388,0.8556701030927835,0.6597938144329897,5.391752577319588,0.3402061855670103,97,0.3402061855670103,31.903225806451612,java,0.0
+5.111219097255348,0.6422018348623854,0.5045871559633027,4.009174311926605,0.4954128440366973,109,0.4954128440366973,29.88888888888889,java,0.0
+5.169436190447377,0.8862275449101796,0.7005988023952097,2.8862275449101795,0.2994011976047904,167,0.2994011976047904,42.2,java,1.0
+4.897236854659848,0.6271186440677966,0.4067796610169492,4.593220338983051,0.5932203389830508,59,0.5932203389830508,29.384615384615383,java,0.0
+5.5494840429514545,0.9004739336492891,0.6966824644549763,6.156398104265403,0.3033175355450237,211,0.3033175355450237,33.016949152542374,java,0.0
+4.316499122181017,0.8412698412698413,0.6031746031746033,2.6507936507936507,0.3968253968253968,63,0.3968253968253968,29.818181818181817,java,1.0
+4.946526678933779,0.5,0.31481481481481477,4.203703703703703,0.6851851851851852,54,0.6851851851851852,48.5,python,1.0
+3.980473809580999,0.9195402298850575,0.7471264367816092,2.528735632183908,0.25287356321839083,87,0.25287356321839083,22.53846153846154,java,0.0
+4.931343145380855,0.8930817610062893,0.6666666666666667,2.0440251572327046,0.3333333333333333,159,0.3333333333333333,25.423076923076923,python,1.0
+5.865359596361463,0.7432432432432432,0.5067567567567568,5.256756756756757,0.49324324324324326,148,0.49324324324324326,34.970588235294116,python,1.0
+5.124006295628484,0.45614035087719296,0.2807017543859649,3.754385964912281,0.7192982456140351,57,0.7192982456140351,38.0,python,0.0
+6.085910006150067,0.8445378151260504,0.6218487394957983,6.142857142857143,0.37815126050420167,238,0.37815126050420167,28.604395604395606,java,1.0
+4.868510971586463,0.7411764705882353,0.5411764705882354,3.2470588235294118,0.4588235294117647,85,0.4588235294117647,28.17391304347826,java,0.0
+4.605408294302115,0.8611111111111112,0.5972222222222222,3.138888888888889,0.4027777777777778,72,0.4027777777777778,33.857142857142854,python,0.0
+4.73995425222158,0.6964285714285714,0.4464285714285714,6.482142857142857,0.5535714285714286,56,0.5535714285714286,33.4375,java,1.0
+4.8675567139411235,0.7701149425287356,0.5517241379310345,4.816091954022989,0.4482758620689655,87,0.4482758620689655,36.13636363636363,java,0.0
+4.246481972040831,0.7454545454545455,0.5272727272727273,4.2727272727272725,0.4727272727272727,55,0.4727272727272727,30.416666666666668,python,0.0
+2.8016094970590277,0.8,0.55,5.0,0.45,20,0.45,28.8,python,1.0
+5.7659039797576215,0.5669291338582677,0.40944881889763785,5.377952755905512,0.5905511811023622,127,0.5905511811023622,35.2962962962963,java,0.0
+4.8369927564722754,0.5978260869565217,0.46739130434782605,5.695652173913044,0.532608695652174,92,0.532608695652174,30.0,python,1.0
+5.650410881834686,0.8103448275862069,0.6091954022988506,4.022988505747127,0.39080459770114945,174,0.39080459770114945,39.65853658536585,java,0.0
+4.33611521033683,0.6363636363636364,0.43181818181818177,4.75,0.5681818181818182,44,0.5681818181818182,34.6,java,1.0
+6.041274356789173,0.8076923076923077,0.657051282051282,3.801282051282051,0.34294871794871795,312,0.34294871794871795,30.096385542168676,java,0.0
+5.289310575466358,0.6428571428571429,0.5,4.190476190476191,0.5,126,0.5,31.615384615384617,python,1.0
+4.700186536808578,0.7741935483870968,0.5,3.532258064516129,0.5,62,0.5,33.76923076923077,python,0.0
+4.180451390892101,0.4444444444444444,0.2592592592592593,4.777777777777778,0.7407407407407407,27,0.7407407407407407,22.25,python,1.0
+4.786213536530154,0.9016393442622951,0.6967213114754098,2.762295081967213,0.30327868852459017,122,0.30327868852459017,29.083333333333332,python,1.0
+5.081486951808688,0.7666666666666667,0.5222222222222221,3.7888888888888888,0.4777777777777778,90,0.4777777777777778,35.8,python,1.0
+4.44849995084568,0.8461538461538461,0.6593406593406593,3.5934065934065935,0.34065934065934067,91,0.34065934065934067,31.045454545454547,python,0.0
+4.7077148025974385,0.37142857142857144,0.19999999999999996,7.228571428571429,0.8,35,0.8,40.888888888888886,java,1.0
+4.51662508214898,0.8837209302325582,0.6627906976744187,3.302325581395349,0.3372093023255814,86,0.3372093023255814,23.25,java,0.0
+4.591236738386586,0.7903225806451613,0.532258064516129,4.193548387096774,0.46774193548387094,62,0.46774193548387094,26.352941176470587,java,1.0
+4.770332655554241,0.7123287671232876,0.5205479452054795,2.5753424657534247,0.4794520547945205,73,0.4794520547945205,14.681818181818182,java,0.0
+4.193064830243576,0.9555555555555556,0.7333333333333334,2.3555555555555556,0.26666666666666666,90,0.26666666666666666,24.68421052631579,python,1.0
+3.935732282211645,0.90625,0.703125,3.1875,0.296875,64,0.296875,34.72727272727273,java,1.0
+5.255459047825728,0.7096774193548387,0.4838709677419355,4.32258064516129,0.5161290322580645,93,0.5161290322580645,47.583333333333336,python,0.0
+5.56295430052689,0.9239543726235742,0.7490494296577948,6.391634980988593,0.2509505703422053,263,0.2509505703422053,51.39655172413793,java,0.0
+4.7624023788690275,0.8846153846153846,0.5897435897435898,3.41025641025641,0.41025641025641024,78,0.41025641025641024,27.95,python,0.0
+4.641341251556263,0.7164179104477612,0.5223880597014925,3.5671641791044775,0.47761194029850745,67,0.47761194029850745,32.69230769230769,java,1.0
+3.9580339180966533,0.9142857142857143,0.7142857142857143,4.014285714285714,0.2857142857142857,70,0.2857142857142857,24.5,python,1.0
+3.821296333684861,0.6071428571428571,0.3928571428571429,4.357142857142857,0.6071428571428571,28,0.6071428571428571,26.5,java,1.0
+5.99080968202175,0.8926553672316384,0.7033898305084746,4.759887005649717,0.2966101694915254,354,0.2966101694915254,40.125,python,0.0
+5.023079206577155,0.8592592592592593,0.674074074074074,2.77037037037037,0.32592592592592595,135,0.32592592592592595,31.88888888888889,java,0.0
+4.427567157116929,0.675,0.4,5.55,0.6,40,0.6,31.4,java,1.0
+4.183636205660773,0.8666666666666667,0.6916666666666667,2.8583333333333334,0.30833333333333335,120,0.30833333333333335,28.38888888888889,python,1.0
+4.6597716630131325,0.6964285714285714,0.4642857142857143,3.892857142857143,0.5357142857142857,56,0.5357142857142857,37.9,java,0.0
+5.011095568495564,0.5454545454545454,0.32727272727272727,5.490909090909091,0.6727272727272727,55,0.6727272727272727,33.93333333333333,java,1.0
+4.07744157582188,0.8085106382978723,0.5531914893617021,2.1702127659574466,0.44680851063829785,47,0.44680851063829785,31.142857142857142,python,0.0
+4.0614821867207755,0.3181818181818182,0.18181818181818177,3.8181818181818183,0.8181818181818182,22,0.8181818181818182,29.75,python,1.0
+4.490765292824625,0.7254901960784313,0.47058823529411764,3.019607843137255,0.5294117647058824,51,0.5294117647058824,22.727272727272727,python,1.0
+3.7695855752178566,0.9612403100775194,0.8372093023255813,2.550387596899225,0.16279069767441862,129,0.16279069767441862,26.833333333333332,java,0.0
+5.113304524601207,0.8288288288288288,0.5945945945945945,3.936936936936937,0.40540540540540543,111,0.40540540540540543,26.9375,java,0.0
+6.683599316096754,0.8333333333333334,0.6434599156118144,5.89662447257384,0.35654008438818563,474,0.35654008438818563,44.127906976744185,python,0.0
+5.742479746741097,0.8826291079812206,0.6572769953051643,7.413145539906103,0.3427230046948357,213,0.3427230046948357,34.11538461538461,java,1.0
+5.684220757851569,0.842809364548495,0.6889632107023411,4.11371237458194,0.3110367892976589,299,0.3110367892976589,42.95348837209303,python,0.0
+3.7248747399889837,0.8125,0.6041666666666667,3.6666666666666665,0.3958333333333333,48,0.3958333333333333,26.857142857142858,python,0.0
+4.438721875540867,0.7083333333333334,0.45833333333333337,3.5625,0.5416666666666666,48,0.5416666666666666,14.647058823529411,java,0.0
+5.052330376272108,0.7604166666666666,0.5416666666666667,5.239583333333333,0.4583333333333333,96,0.4583333333333333,37.76190476190476,java,0.0
+4.039211536948156,0.5161290322580645,0.3548387096774194,3.3870967741935485,0.6451612903225806,31,0.6451612903225806,12.0,java,0.0
+6.058508660286119,0.8492307692307692,0.6830769230769231,6.8246153846153845,0.3169230769230769,325,0.3169230769230769,40.19047619047619,java,1.0
+4.85654598061365,0.8559322033898306,0.6440677966101696,3.610169491525424,0.3559322033898305,118,0.3559322033898305,22.06896551724138,java,0.0
+4.384791749673746,0.8899082568807339,0.7155963302752293,2.6055045871559632,0.28440366972477066,109,0.28440366972477066,33.44444444444444,java,1.0
+4.2115044085684765,0.631578947368421,0.42105263157894735,4.473684210526316,0.5789473684210527,38,0.5789473684210527,31.77777777777778,python,1.0
+6.445861001406715,0.7741935483870968,0.5443548387096775,5.548387096774194,0.45564516129032256,248,0.45564516129032256,36.03508771929825,python,0.0
+4.060861683988659,0.8043478260869565,0.5652173913043479,3.0,0.43478260869565216,46,0.43478260869565216,28.77777777777778,python,1.0
+4.770281067729926,0.6226415094339622,0.39622641509433965,4.962264150943396,0.6037735849056604,53,0.6037735849056604,29.642857142857142,java,1.0
+4.310866879118227,0.8829787234042553,0.6914893617021276,4.340425531914893,0.30851063829787234,94,0.30851063829787234,30.791666666666668,python,0.0
+4.912733576964344,0.5636363636363636,0.36363636363636365,3.4545454545454546,0.6363636363636364,55,0.6363636363636364,24.705882352941178,python,1.0
+3.76062054430996,0.6666666666666666,0.48484848484848486,3.8484848484848486,0.5151515151515151,33,0.5151515151515151,27.11111111111111,python,1.0
+5.092277224236287,0.6621621621621622,0.43243243243243246,3.7837837837837838,0.5675675675675675,74,0.5675675675675675,29.5625,java,0.0
+3.5,0.5,0.25,3.125,0.75,16,0.75,19.75,python,1.0
+4.62765192370225,0.952,0.736,2.472,0.264,125,0.264,35.714285714285715,python,0.0
+4.516221717727311,0.8026315789473685,0.6052631578947368,5.197368421052632,0.39473684210526316,76,0.39473684210526316,27.72222222222222,python,0.0
+3.937751955408587,0.6451612903225806,0.4193548387096774,4.193548387096774,0.5806451612903226,31,0.5806451612903226,29.25,python,0.0
+5.000896232556761,0.8045977011494253,0.5402298850574713,3.6091954022988504,0.45977011494252873,87,0.45977011494252873,33.0,python,1.0
+6.038322688621861,0.827906976744186,0.6046511627906976,6.120930232558139,0.3953488372093023,215,0.3953488372093023,49.08108108108108,python,0.0
+4.216152190789069,0.8163265306122449,0.5510204081632653,3.5714285714285716,0.4489795918367347,49,0.4489795918367347,22.263157894736842,java,0.0
+4.667187378614269,0.8235294117647058,0.5882352941176471,4.0588235294117645,0.4117647058823529,85,0.4117647058823529,32.857142857142854,java,0.0
+5.070332970220562,0.8478260869565217,0.6594202898550725,4.297101449275362,0.34057971014492755,138,0.34057971014492755,30.34285714285714,java,0.0
+4.2366344310623845,0.9058823529411765,0.7058823529411764,3.8941176470588235,0.29411764705882354,85,0.29411764705882354,35.0,java,1.0
+4.062382847947341,0.5454545454545454,0.36363636363636365,2.393939393939394,0.6363636363636364,33,0.6363636363636364,24.333333333333332,python,1.0
+4.477226049206228,0.6851851851851852,0.4629629629629629,3.8703703703703702,0.5370370370370371,54,0.5370370370370371,37.0,java,1.0
+4.189125478116705,0.6521739130434783,0.4782608695652174,5.217391304347826,0.5217391304347826,46,0.5217391304347826,39.77777777777778,java,1.0
+3.6818808028034025,0.47368421052631576,0.26315789473684215,3.3157894736842106,0.7368421052631579,19,0.7368421052631579,22.75,python,1.0
+4.183768266288633,0.78125,0.609375,1.96875,0.390625,64,0.390625,23.25,python,1.0
+4.063202513215852,0.7368421052631579,0.4736842105263158,3.1052631578947367,0.5263157894736842,38,0.5263157894736842,23.77777777777778,python,1.0
+5.240360496391271,0.7901234567901234,0.4814814814814815,4.592592592592593,0.5185185185185185,81,0.5185185185185185,35.0,java,0.0
+2.9870900416203345,0.7619047619047619,0.5238095238095238,4.333333333333333,0.47619047619047616,21,0.47619047619047616,18.444444444444443,python,1.0
+4.796230935589923,0.917910447761194,0.7238805970149254,3.1194029850746268,0.27611940298507465,134,0.27611940298507465,33.964285714285715,java,0.0
+4.921133626426239,0.6779661016949152,0.4067796610169492,4.576271186440678,0.5932203389830508,59,0.5932203389830508,37.9,python,0.0
+5.223430137239087,0.9595375722543352,0.838150289017341,3.3757225433526012,0.16184971098265896,346,0.16184971098265896,26.730337078651687,java,0.0
+5.035549713367317,0.5454545454545454,0.32727272727272727,3.7636363636363637,0.6727272727272727,55,0.6727272727272727,25.272727272727273,python,0.0
+5.404449012034224,0.7906976744186046,0.5658914728682171,5.658914728682171,0.43410852713178294,129,0.43410852713178294,36.758620689655174,java,0.0
+4.629412138272088,0.4418604651162791,0.3023255813953488,3.8372093023255816,0.6976744186046512,43,0.6976744186046512,27.7,python,0.0
+4.816777664828826,0.9296875,0.71875,4.7109375,0.28125,128,0.28125,37.078947368421055,java,0.0
+5.5887549870381,0.8138297872340425,0.6276595744680851,4.388297872340425,0.3723404255319149,188,0.3723404255319149,50.24,python,0.0
+4.453150309616965,0.8518518518518519,0.7185185185185186,2.4074074074074074,0.2814814814814815,135,0.2814814814814815,14.0,java,0.0
+5.547057692620618,0.9096385542168675,0.7710843373493976,4.421686746987952,0.2289156626506024,332,0.2289156626506024,31.275862068965516,java,0.0
+4.627801780853476,0.9,0.6555555555555556,5.055555555555555,0.34444444444444444,90,0.34444444444444444,34.04761904761905,java,0.0
+4.747191429566854,0.5333333333333333,0.3111111111111111,5.466666666666667,0.6888888888888889,45,0.6888888888888889,36.44444444444444,python,0.0
+5.524469760991255,0.6788990825688074,0.45871559633027525,4.91743119266055,0.5412844036697247,109,0.5412844036697247,53.38461538461539,python,0.0
+5.347434988263512,0.4444444444444444,0.29166666666666663,4.0,0.7083333333333334,72,0.7083333333333334,45.625,python,0.0
+4.074705922236444,0.847457627118644,0.6271186440677966,3.6779661016949152,0.3728813559322034,59,0.3728813559322034,23.1,python,1.0
+4.41506101220307,0.43333333333333335,0.23333333333333328,4.833333333333333,0.7666666666666667,30,0.7666666666666667,17.333333333333332,python,0.0
+5.489282954487391,0.7518796992481203,0.5488721804511278,6.7443609022556394,0.45112781954887216,133,0.45112781954887216,41.4,java,0.0
+4.3199650731621695,0.8584070796460177,0.7256637168141593,3.3628318584070795,0.2743362831858407,113,0.2743362831858407,26.551724137931036,python,0.0
+5.948852612281259,0.8584905660377359,0.6886792452830188,6.908805031446541,0.3113207547169811,318,0.3113207547169811,39.423529411764704,java,1.0
+4.251365544851387,0.957983193277311,0.773109243697479,2.302521008403361,0.226890756302521,119,0.226890756302521,25.958333333333332,python,1.0
+4.966534113288597,0.6,0.41538461538461535,3.753846153846154,0.5846153846153846,65,0.5846153846153846,34.45454545454545,python,1.0
+4.216421766792934,0.6808510638297872,0.4893617021276596,3.4468085106382977,0.5106382978723404,47,0.5106382978723404,25.53846153846154,python,1.0
+5.854429565880467,0.47619047619047616,0.32380952380952377,5.79047619047619,0.6761904761904762,105,0.6761904761904762,32.892857142857146,python,1.0
+5.542084533777365,0.8870967741935484,0.7217741935483871,4.866935483870968,0.2782258064516129,248,0.2782258064516129,38.97727272727273,java,0.0
+3.8370959941879037,0.7560975609756098,0.5609756097560976,2.5853658536585367,0.43902439024390244,41,0.43902439024390244,23.8,python,1.0
+5.510000244558343,0.8188405797101449,0.572463768115942,5.884057971014493,0.427536231884058,138,0.427536231884058,31.333333333333332,java,0.0
+3.9171736218037974,0.9090909090909091,0.7121212121212122,2.878787878787879,0.2878787878787879,66,0.2878787878787879,30.307692307692307,java,1.0
+4.955593373471456,0.9126984126984127,0.6904761904761905,2.7222222222222223,0.30952380952380953,126,0.30952380952380953,31.208333333333332,python,1.0
+4.123966016311361,0.7407407407407407,0.537037037037037,3.074074074074074,0.46296296296296297,54,0.46296296296296297,29.9,python,0.0
+5.280191443467718,0.7971014492753623,0.6086956521739131,4.173913043478261,0.391304347826087,138,0.391304347826087,21.972972972972972,java,0.0
+4.616874605956221,0.47058823529411764,0.23529411764705888,5.147058823529412,0.7647058823529411,34,0.7647058823529411,33.142857142857146,python,1.0
+4.128326967635621,0.8043478260869565,0.5434782608695652,4.108695652173913,0.45652173913043476,46,0.45652173913043476,30.363636363636363,java,1.0
+4.589416279148799,0.8409090909090909,0.6363636363636364,2.715909090909091,0.36363636363636365,88,0.36363636363636365,17.785714285714285,java,0.0
+4.935672972133341,0.9946236559139785,0.8064516129032258,4.091397849462366,0.1935483870967742,186,0.1935483870967742,31.8,java,0.0
+4.76880107599223,0.8446601941747572,0.6310679611650485,3.8446601941747574,0.36893203883495146,103,0.36893203883495146,52.72727272727273,python,0.0
+4.1699823363507,0.8478260869565217,0.5434782608695652,5.456521739130435,0.45652173913043476,46,0.45652173913043476,33.45454545454545,java,1.0
+4.017917900762096,0.7058823529411765,0.4411764705882353,4.117647058823529,0.5588235294117647,34,0.5588235294117647,30.571428571428573,python,0.0
+4.6586089678214995,0.896,0.704,2.696,0.296,125,0.296,31.40740740740741,python,0.0
+4.643160624658742,0.7258064516129032,0.5,3.3870967741935485,0.5,62,0.5,45.30769230769231,java,0.0
+4.813012081410429,0.7857142857142857,0.5285714285714286,4.257142857142857,0.4714285714285714,70,0.4714285714285714,31.647058823529413,python,1.0
+4.60865917212318,0.9142857142857143,0.6952380952380952,2.6952380952380954,0.3047619047619048,105,0.3047619047619048,28.5,python,1.0
+4.713243393050852,0.9347826086956522,0.7536231884057971,2.152173913043478,0.2463768115942029,138,0.2463768115942029,33.76190476190476,python,0.0
+4.619054056131191,0.734375,0.515625,3.296875,0.484375,64,0.484375,28.066666666666666,java,0.0
+4.825245826488317,0.7654320987654321,0.5308641975308642,4.580246913580247,0.4691358024691358,81,0.4691358024691358,33.9,java,0.0
+6.527235468349858,0.8012232415902141,0.6085626911314985,5.792048929663609,0.39143730886850153,327,0.39143730886850153,42.516129032258064,python,0.0
+5.59181229230442,0.7133333333333334,0.5333333333333333,4.233333333333333,0.4666666666666667,150,0.4666666666666667,33.55172413793103,python,0.0
+4.0006374384002905,0.8421052631578947,0.631578947368421,2.491228070175439,0.3684210526315789,57,0.3684210526315789,33.5,python,1.0
+3.7461571189724303,0.9545454545454546,0.6590909090909092,4.25,0.3409090909090909,44,0.3409090909090909,28.833333333333332,python,1.0
+5.745512973329542,0.8275862068965517,0.5517241379310345,6.027586206896552,0.4482758620689655,145,0.4482758620689655,34.648648648648646,java,0.0
+5.490266518243658,0.8412698412698413,0.6243386243386244,5.121693121693122,0.37566137566137564,189,0.37566137566137564,40.166666666666664,java,0.0
+4.496115365169272,0.6875,0.4375,4.729166666666667,0.5625,48,0.5625,43.285714285714285,python,1.0
+5.505436888624101,0.945273631840796,0.7164179104477613,3.855721393034826,0.2835820895522388,201,0.2835820895522388,35.58139534883721,java,0.0
+4.795079572399797,0.8155339805825242,0.6116504854368932,4.621359223300971,0.3883495145631068,103,0.3883495145631068,37.54545454545455,java,0.0
+5.595002345872928,0.7768595041322314,0.5206611570247934,6.099173553719008,0.4793388429752066,121,0.4793388429752066,32.029411764705884,java,0.0
+5.887085632611005,0.93359375,0.68359375,6.109375,0.31640625,256,0.31640625,31.675675675675677,java,0.0
+4.809267348202161,0.7971014492753623,0.5217391304347826,3.8840579710144927,0.4782608695652174,69,0.4782608695652174,38.15384615384615,python,0.0
+4.893291230757242,0.5737704918032787,0.39344262295081966,4.311475409836065,0.6065573770491803,61,0.6065573770491803,33.7,python,0.0
+4.346926022805004,0.8,0.5454545454545454,4.072727272727272,0.45454545454545453,55,0.45454545454545453,29.866666666666667,java,0.0
+4.476474118254376,0.7857142857142857,0.5178571428571428,6.660714285714286,0.48214285714285715,56,0.48214285714285715,39.92307692307692,java,1.0
+3.9809099797950775,0.9154929577464789,0.704225352112676,4.422535211267606,0.29577464788732394,71,0.29577464788732394,30.0,java,1.0
+5.272508205443481,0.5967741935483871,0.32258064516129037,6.193548387096774,0.6774193548387096,62,0.6774193548387096,44.166666666666664,java,1.0
+5.1902361166256314,0.9676113360323887,0.7854251012145749,4.611336032388664,0.2145748987854251,247,0.2145748987854251,28.724137931034484,java,1.0
+4.350489295854283,0.7333333333333333,0.55,3.183333333333333,0.45,60,0.45,33.8,java,1.0
+5.0009504388098796,0.7927927927927928,0.5765765765765766,4.5225225225225225,0.42342342342342343,111,0.42342342342342343,35.142857142857146,java,0.0
+5.7769401731183265,0.9180327868852459,0.6284153005464481,5.573770491803279,0.37158469945355194,183,0.37158469945355194,28.34375,java,0.0
+3.876189125313799,0.8970588235294118,0.6911764705882353,2.073529411764706,0.3088235294117647,68,0.3088235294117647,27.9,java,1.0
+4.032303242743952,0.4444444444444444,0.2962962962962963,3.888888888888889,0.7037037037037037,27,0.7037037037037037,26.5,python,0.0
+3.625,0.375,0.1875,6.625,0.8125,16,0.8125,27.8,python,1.0
+4.4210998675908275,0.9540816326530612,0.826530612244898,2.4846938775510203,0.17346938775510204,196,0.17346938775510204,35.06896551724138,python,0.0
+4.839775539645508,0.3157894736842105,0.1842105263157895,5.157894736842105,0.8157894736842105,38,0.8157894736842105,37.0,python,1.0
+5.842252978929465,0.8711484593837535,0.7254901960784313,6.647058823529412,0.27450980392156865,357,0.27450980392156865,42.16470588235294,java,1.0
+4.166886761135438,0.78,0.56,3.2,0.44,50,0.44,29.916666666666668,python,0.0
+5.395683189409616,0.8947368421052632,0.6947368421052631,5.515789473684211,0.30526315789473685,190,0.30526315789473685,29.24590163934426,java,1.0

basemodel/train_randomforest.py

@@ -0,0 +1,44 @@
+import pandas as pd
+import joblib
+from sklearn.ensemble import RandomForestClassifier
+from sklearn.metrics import accuracy_score, classification_report
+
+
+train_df = pd.read_csv("basemodel/train_features.csv")
+val_df = pd.read_csv("basemodel/val_features.csv")
+
+
+drop_cols = ["Label"]
+if "language" in train_df.columns:
+    drop_cols.append("language")
+
+X_train = train_df.drop(columns=drop_cols)
+y_train = train_df["Label"]
+
+X_val = val_df.drop(columns=drop_cols)
+y_val = val_df["Label"]
+
+
+rf = RandomForestClassifier(
+    n_estimators=200,
+    max_depth=8,
+    min_samples_split=5,
+    min_samples_leaf=3,
+    random_state=42,
+    class_weight="balanced"
+)
+
+rf.fit(X_train, y_train)
+
+
+val_preds = rf.predict(X_val)
+
+accuracy = accuracy_score(y_val, val_preds)
+print("\nValidation Accuracy:", round(accuracy, 4))
+
+print("\nValidation Classification Report:\n")
+print(classification_report(y_val, val_preds, target_names=["Human", "AI"]))
+
+
+joblib.dump(rf, "basemodel/random_forest_baseline.pkl")
+print("\n✅ Random Forest baseline model saved to basemodel/random_forest_baseline.pkl")

basemodel/val_features.csv

@@ -0,0 +1,75 @@
+entropy,burstiness,repetition_ratio,avg_token_length,vocab_richness,num_tokens,unique_token_ratio,avg_line_length,language,Label
+3.616348566075164,0.42105263157894735,0.26315789473684215,3.4210526315789473,0.7368421052631579,19,0.7368421052631579,25.25,python,1.0
+5.225116096801971,0.9219512195121952,0.7365853658536585,4.84390243902439,0.2634146341463415,205,0.2634146341463415,31.340425531914892,java,0.0
+5.858683407719383,0.5053763440860215,0.30107526881720426,6.021505376344086,0.6989247311827957,93,0.6989247311827957,30.59259259259259,python,1.0
+3.4687406100460665,0.918918918918919,0.6486486486486487,2.4324324324324325,0.35135135135135137,37,0.35135135135135137,40.333333333333336,python,1.0
+4.726474118254375,0.75,0.4642857142857143,6.339285714285714,0.5357142857142857,56,0.5357142857142857,33.25,java,1.0
+5.127518057233353,0.9186602870813397,0.7416267942583732,4.779904306220096,0.2583732057416268,209,0.2583732057416268,42.84848484848485,java,0.0
+3.385592490318422,0.8181818181818182,0.6909090909090909,2.6,0.3090909090909091,55,0.3090909090909091,23.75,python,1.0
+5.303371529748239,0.8614457831325302,0.6566265060240963,4.445783132530121,0.3433734939759036,166,0.3433734939759036,19.36,java,0.0
+4.32307418942857,0.72,0.5,4.04,0.5,50,0.5,23.647058823529413,python,1.0
+4.646104113627779,0.6666666666666666,0.4444444444444444,3.925925925925926,0.5555555555555556,54,0.5555555555555556,17.8125,java,0.0
+4.824043435284101,0.7971014492753623,0.5217391304347826,3.9130434782608696,0.4782608695652174,69,0.4782608695652174,26.157894736842106,python,0.0
+4.133660689688186,0.44,0.24,6.0,0.76,25,0.76,28.625,java,1.0
+3.977232016069148,0.8854166666666666,0.7395833333333333,2.7395833333333335,0.2604166666666667,96,0.2604166666666667,28.4375,python,1.0
+4.418945246473782,0.8461538461538461,0.641025641025641,3.1153846153846154,0.358974358974359,78,0.358974358974359,33.78947368421053,java,0.0
+3.795088586397732,0.4782608695652174,0.30434782608695654,3.5652173913043477,0.6956521739130435,23,0.6956521739130435,21.166666666666668,python,1.0
+4.665893393157604,0.8735632183908046,0.632183908045977,2.4942528735632186,0.367816091954023,87,0.367816091954023,30.833333333333332,java,1.0
+4.888279432092592,0.5636363636363636,0.36363636363636365,4.5636363636363635,0.6363636363636364,55,0.6363636363636364,31.7,python,0.0
+4.43887358554625,0.7111111111111111,0.4444444444444444,4.488888888888889,0.5555555555555556,45,0.5555555555555556,27.533333333333335,java,0.0
+4.956005194695069,0.7636363636363637,0.5909090909090908,3.963636363636364,0.4090909090909091,110,0.4090909090909091,22.925925925925927,java,0.0
+4.364611584126849,0.8990825688073395,0.7201834862385321,3.86697247706422,0.2798165137614679,218,0.2798165137614679,36.03125,python,0.0
+5.1310450521201885,0.6923076923076923,0.4945054945054945,3.8241758241758244,0.5054945054945055,91,0.5054945054945055,43.083333333333336,python,0.0
+4.157269582127642,0.9487179487179487,0.7863247863247863,1.735042735042735,0.21367521367521367,117,0.21367521367521367,36.583333333333336,python,1.0
+6.1032071909268755,0.7524271844660194,0.558252427184466,4.344660194174757,0.441747572815534,206,0.441747572815534,42.59375,python,0.0
+5.489504669651612,0.9067796610169492,0.7203389830508475,3.983050847457627,0.2796610169491525,236,0.2796610169491525,46.03225806451613,java,0.0
+5.5429756373358305,0.6216216216216216,0.4504504504504504,3.810810810810811,0.5495495495495496,111,0.5495495495495496,46.0,python,0.0
+5.597773058187228,0.7096774193548387,0.4193548387096774,4.548387096774194,0.5806451612903226,93,0.5806451612903226,32.27777777777778,python,1.0
+4.288320189106,0.7916666666666666,0.5208333333333333,4.229166666666667,0.4791666666666667,48,0.4791666666666667,34.0,python,0.0
+5.117143809076922,0.8490566037735849,0.5849056603773585,3.943396226415094,0.41509433962264153,106,0.41509433962264153,39.4,python,0.0
+4.3341837197791895,0.8,0.44999999999999996,7.05,0.55,40,0.55,41.666666666666664,java,1.0
+4.7319988769018755,0.8416666666666667,0.675,3.091666666666667,0.325,120,0.325,15.944444444444445,java,0.0
+4.2628148954723555,0.55,0.4,2.825,0.6,40,0.6,31.285714285714285,python,0.0
+3.25,0.875,0.625,3.6875,0.375,32,0.375,41.4,java,1.0
+5.086021506160257,0.7008547008547008,0.5470085470085471,4.239316239316239,0.452991452991453,117,0.452991452991453,44.125,python,0.0
+4.488525294897748,0.6923076923076923,0.46153846153846156,3.9038461538461537,0.5384615384615384,52,0.5384615384615384,30.181818181818183,java,1.0
+5.273684376262023,0.6111111111111112,0.375,4.819444444444445,0.625,72,0.625,42.72727272727273,python,0.0
+4.047055675509121,0.8627450980392157,0.607843137254902,4.372549019607843,0.39215686274509803,51,0.39215686274509803,28.428571428571427,python,1.0
+4.6442547530701885,0.5370370370370371,0.38888888888888884,5.12962962962963,0.6111111111111112,54,0.6111111111111112,34.7,java,0.0
+4.423251796980337,0.32142857142857145,0.1785714285714286,2.642857142857143,0.8214285714285714,28,0.8214285714285714,30.0,python,1.0
+5.498927847038898,0.5842696629213483,0.3820224719101124,6.617977528089888,0.6179775280898876,89,0.6179775280898876,41.421052631578945,python,1.0
+4.376857192109841,0.8625,0.65,3.4125,0.35,80,0.35,29.764705882352942,java,0.0
+4.600176636857892,0.7894736842105263,0.5087719298245614,3.6842105263157894,0.49122807017543857,57,0.49122807017543857,30.5,python,1.0
+6.213552119915352,0.8066037735849056,0.5613207547169812,5.485849056603773,0.4386792452830189,212,0.4386792452830189,39.6,python,0.0
+4.024059741690056,0.7727272727272727,0.6060606060606061,3.5454545454545454,0.3939393939393939,66,0.3939393939393939,30.46153846153846,python,1.0
+4.14853573331382,0.7619047619047619,0.5,7.023809523809524,0.5,42,0.5,26.266666666666666,java,0.0
+4.2057539822273915,0.8507462686567164,0.6417910447761195,3.5522388059701493,0.3582089552238806,67,0.3582089552238806,30.352941176470587,python,0.0
+4.860498585099573,0.8737864077669902,0.6504854368932038,4.097087378640777,0.34951456310679613,103,0.34951456310679613,32.61538461538461,java,0.0
+4.661648023061636,0.7631578947368421,0.5526315789473684,3.776315789473684,0.4473684210526316,76,0.4473684210526316,32.05555555555556,java,0.0
+5.405165733375986,0.9311926605504587,0.7339449541284404,5.426605504587156,0.26605504587155965,218,0.26605504587155965,33.0,java,0.0
+3.238901256602631,0.46153846153846156,0.23076923076923073,4.538461538461538,0.7692307692307693,13,0.7692307692307693,26.0,python,1.0
+3.964395529526285,0.9512195121951219,0.5853658536585367,4.170731707317073,0.4146341463414634,41,0.4146341463414634,27.1,java,1.0
+4.091031702849688,0.8170731707317073,0.6707317073170731,3.8780487804878048,0.32926829268292684,82,0.32926829268292684,31.833333333333332,python,0.0
+3.6596769481795466,0.8571428571428571,0.6190476190476191,4.976190476190476,0.38095238095238093,42,0.38095238095238093,24.533333333333335,java,1.0
+5.249145047994913,0.6875,0.5178571428571428,3.8125,0.48214285714285715,112,0.48214285714285715,46.5,python,0.0
+4.456796057875661,0.7761194029850746,0.582089552238806,2.283582089552239,0.417910447761194,67,0.417910447761194,25.545454545454547,java,1.0
+5.669233537529181,0.8121827411167513,0.6142131979695431,4.3908629441624365,0.38578680203045684,197,0.38578680203045684,49.56,python,0.0
+4.923879361943277,0.9572649572649573,0.6923076923076923,4.205128205128205,0.3076923076923077,117,0.3076923076923077,31.321428571428573,python,0.0
+4.028579035079149,0.9841269841269841,0.7142857142857143,3.8253968253968256,0.2857142857142857,63,0.2857142857142857,32.666666666666664,python,1.0
+4.931270849848961,0.6,0.36363636363636365,5.763636363636364,0.6363636363636364,55,0.6363636363636364,29.8125,java,1.0
+4.987993396430413,0.8682170542635659,0.6744186046511628,3.317829457364341,0.32558139534883723,129,0.32558139534883723,33.02777777777778,java,0.0
+4.895499903544709,0.9303482587064676,0.7661691542288558,3.4776119402985075,0.23383084577114427,201,0.23383084577114427,23.032258064516128,java,0.0
+4.297689964750703,0.9158878504672897,0.7476635514018692,3.532710280373832,0.2523364485981308,107,0.2523364485981308,31.107142857142858,java,0.0
+4.531740514375559,0.8,0.509090909090909,5.9818181818181815,0.4909090909090909,55,0.4909090909090909,30.25,java,1.0
+4.894400915540714,0.7763157894736842,0.5263157894736843,3.3157894736842106,0.47368421052631576,76,0.47368421052631576,29.952380952380953,python,0.0
+5.754451229096238,0.7107438016528925,0.4628099173553719,5.603305785123967,0.5371900826446281,121,0.5371900826446281,34.03703703703704,java,0.0
+3.5141016549301787,0.6944444444444444,0.5555555555555556,3.0277777777777777,0.4444444444444444,36,0.4444444444444444,33.6,python,1.0
+5.4389854001411315,0.7350427350427351,0.5213675213675213,5.170940170940171,0.47863247863247865,117,0.47863247863247865,43.72727272727273,java,0.0
+6.500985818092827,0.8327974276527331,0.6109324758842444,6.864951768488746,0.3890675241157556,311,0.3890675241157556,38.79775280898876,java,1.0
+3.9219280948873627,0.4,0.19999999999999996,3.85,0.8,20,0.8,20.5,python,1.0
+4.594928174344042,0.9072164948453608,0.6804123711340206,2.6804123711340204,0.31958762886597936,97,0.31958762886597936,35.5,java,1.0
+6.3478254862783485,0.7662337662337663,0.5454545454545454,6.783549783549783,0.45454545454545453,231,0.45454545454545453,33.81818181818182,java,1.0
+4.777798184597433,0.8349514563106796,0.6310679611650485,4.475728155339806,0.36893203883495146,103,0.36893203883495146,30.413793103448278,python,0.0
+4.955482172039427,0.7422680412371134,0.5257731958762887,5.556701030927835,0.4742268041237113,97,0.4742268041237113,29.88,python,0.0
+3.963119345888566,0.7291666666666666,0.5416666666666667,3.8333333333333335,0.4583333333333333,48,0.4583333333333333,29.2,python,1.0
+4.981129381113054,0.7857142857142857,0.5892857142857143,3.8214285714285716,0.4107142857142857,112,0.4107142857142857,47.69230769230769,python,0.0

classifier/inference.py

@@ -0,0 +1,130 @@
+import sys
+import os
+
+PROJECT_ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+sys.path.append(PROJECT_ROOT)
+
+import numpy as np
+import pandas as pd
+import xgboost as xgb
+
+# -------------------------------
+# IMPORT FEATURE EXTRACTORS
+# -------------------------------
+from featureextraction.step1_statistical_extraction.step1_statistical_extraction import extract_features as extract_stat
+from featureextraction.step2_ast_extraction.step2_ast_extraction import extract_ast_features
+from featureextraction.step3_stylometry_extraction.step3_stylometry_extraction import extract_stylometry_features
+from featureextraction.semantic_features.unixcoder_embedding import get_unixcoder_embedding
+
+# XAI modules
+from xai.shaplayer import shap_explain
+from xai.grouping import group_shap_explanations
+from xai.text_explainer import generate_text_explanation
+
+# -------------------------------
+# LOAD MODEL
+# -------------------------------
+model = xgb.XGBClassifier()
+model.load_model("classifier/xgboost_final_model.json")
+
+# -------------------------------
+# LANGUAGE ONE-HOT
+# -------------------------------
+def encode_language(language):
+    language = language.lower()
+    if language == "python":
+        return np.array([1, 0])
+    elif language == "java":
+        return np.array([0, 1])
+    else:
+        raise ValueError("Language must be python or java")
+
+# -------------------------------
+# BUILD FEATURES FROM CODE
+# -------------------------------
+def build_features_from_code(code, language):
+
+    df = pd.DataFrame({
+        "normalized_code": [code],
+        "Language": [language]
+    })
+
+    stat_df = extract_stat(df)
+    ast_df = extract_ast_features(df)
+    style_df = extract_stylometry_features(df)
+
+    X_stat = stat_df.drop(columns=["language"]).values.flatten()
+    X_ast = ast_df.values.flatten()
+    X_style = style_df.values.flatten()
+    X_lang = encode_language(language)
+    X_sem = get_unixcoder_embedding(code)
+
+    X_final = np.hstack([
+        X_stat,
+        X_ast,
+        X_style,
+        X_lang,
+        X_sem
+    ]).reshape(1, -1)
+
+    return X_final
+
+# -------------------------------
+# BASIC PREDICT FUNCTION
+# -------------------------------
+def predict_from_features(X_final):
+    prediction = model.predict(X_final)[0]
+    probability = model.predict_proba(X_final)[0][1]
+    label_name = "AI" if prediction == 1 else "Human"
+    return label_name, probability
+
+# -------------------------------
+# INTERACTIVE CLI
+# -------------------------------
+if __name__ == "__main__":
+
+    print("\n======================================")
+    print("  AI vs Human Code Classification")
+    print("======================================")
+
+    language = input("Choose language (python/java): ").strip().lower()
+
+    print("\nPaste your code below.")
+    print("Type 'END' on a new line when finished.\n")
+
+    lines = []
+    while True:
+        line = input()
+        if line.strip() == "END":
+            break
+        lines.append(line)
+
+    code_input = "\n".join(lines)
+
+    # build features
+    X_final = build_features_from_code(code_input, language)
+
+    # predict
+    label, prob = predict_from_features(X_final)
+
+    # shap
+    shap_result = shap_explain(model, X_final)
+
+    # grouping
+    grouped = group_shap_explanations(shap_result)
+
+    # text explanation
+    text_reason = generate_text_explanation(grouped, label, prob)
+
+    print("\n========== RESULT ==========")
+    print("Prediction :", label)
+    print("Confidence :", prob)
+
+    print("\nTop SHAP features:")
+    for e in shap_result:
+        print(f"Feature {e['feature_index']} → {e['impact']} ({e['pushes_toward']})")
+
+    print("\nGrouped SHAP importance:", grouped)
+
+    print("\nExplanation:\n")
+    print(text_reason)

classifier/test_xgboost.py

@@ -0,0 +1,28 @@
+import numpy as np
+import xgboost as xgb
+from sklearn.metrics import accuracy_score, classification_report, confusion_matrix
+
+X_test = np.load("featureextraction/final_features/test_X.npy")
+y_test = np.load("featureextraction/final_features/test_y.npy")
+
+print("Test shape:", X_test.shape)
+
+model = xgb.XGBClassifier()
+model.load_model("classifier/xgboost_final_model.json")
+
+y_pred = model.predict(X_test)
+
+print("\nTEST SET RESULTS (FINAL MODEL)\n")
+print("Accuracy:", accuracy_score(y_test, y_pred))
+
+print("\nClassification Report:\n")
+print(
+    classification_report(
+        y_test,
+        y_pred,
+        target_names=["Human", "AI"]
+    )
+)
+
+print("\nConfusion Matrix:\n")
+print(confusion_matrix(y_test, y_pred))

classifier/train_xgboost.py

@@ -0,0 +1,68 @@
+import numpy as np
+import xgboost as xgb
+from sklearn.metrics import accuracy_score, classification_report
+import joblib
+import os
+
+PROJECT_ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+
+FEATURE_DIR = os.path.join(
+    PROJECT_ROOT,
+    "featureextraction",
+    "final_features"
+)
+
+X_train = np.load(
+    os.path.join(FEATURE_DIR, "train_X.npy"),
+    allow_pickle=True
+)
+y_train = np.load(
+    os.path.join(FEATURE_DIR, "train_y.npy"),
+    allow_pickle=True
+)
+
+X_val = np.load(
+    os.path.join(FEATURE_DIR, "val_X.npy"),
+    allow_pickle=True
+)
+y_val = np.load(
+    os.path.join(FEATURE_DIR, "val_y.npy"),
+    allow_pickle=True
+)
+
+print("Train shape:", X_train.shape)
+print("Validation shape:", X_val.shape)
+print("Feature dtype:", X_train.dtype)
+
+model = xgb.XGBClassifier(
+    n_estimators=300,
+    max_depth=6,
+    learning_rate=0.05,
+    subsample=0.8,
+    colsample_bytree=0.8,
+    objective="binary:logistic",
+    eval_metric="logloss",
+    random_state=42
+)
+
+model.fit(X_train, y_train)
+
+val_preds = model.predict(X_val)
+
+print("\nVALIDATION RESULTS\n")
+print("Accuracy:", accuracy_score(y_val, val_preds))
+print(
+    classification_report(
+        y_val,
+        val_preds,
+        target_names=["Human", "AI"]
+    )
+)
+
+MODEL_DIR = os.path.join(PROJECT_ROOT, "models")
+os.makedirs(MODEL_DIR, exist_ok=True)
+
+model.save_model(os.path.join(MODEL_DIR, "xgboost_final_model.json"))
+joblib.dump(model, os.path.join(MODEL_DIR, "xgboost_final_model.pkl"))
+
+print("\n XGBoost model saved successfully")

classifier/tune_xgboost.py

@@ -0,0 +1,72 @@
+import numpy as np
+import xgboost as xgb
+from sklearn.metrics import accuracy_score, classification_report
+from sklearn.model_selection import ParameterGrid
+import joblib
+
+X_train = np.load("featureextraction/final_features/train_X.npy")
+y_train = np.load("featureextraction/final_features/train_y.npy")
+
+X_val = np.load("featureextraction/final_features/val_X.npy")
+y_val = np.load("featureextraction/final_features/val_y.npy")
+
+print("Train shape:", X_train.shape)
+print("Val shape:", X_val.shape)
+
+param_grid = {
+    "n_estimators": [300, 500],
+    "max_depth": [4, 6, 8],
+    "learning_rate": [0.03, 0.05, 0.1],
+    "subsample": [0.8, 1.0],
+    "colsample_bytree": [0.8, 1.0]
+}
+
+best_acc = 0.0
+best_model = None
+best_params = None
+
+print("\n Starting XGBoost Hyperparameter Tuning...\n")
+
+for idx, params in enumerate(ParameterGrid(param_grid), start=1):
+    print(f"Trial {idx}: {params}")
+
+    model = xgb.XGBClassifier(
+        objective="binary:logistic",
+        eval_metric="logloss",
+        random_state=42,
+        **params
+    )
+
+    model.fit(X_train, y_train)
+
+    val_preds = model.predict(X_val)
+    acc = accuracy_score(y_val, val_preds)
+
+    print("Validation Accuracy:", round(acc, 4))
+
+    if acc > best_acc:
+        best_acc = acc
+        best_model = model
+        best_params = params
+        print("New best model found")
+
+    print("-" * 60)
+
+print("\n🏆 BEST VALIDATION ACCURACY:", round(best_acc, 4))
+print("🏆 BEST PARAMETERS:", best_params)
+
+final_preds = best_model.predict(X_val)
+
+print("\nFINAL VALIDATION REPORT\n")
+print(
+    classification_report(
+        y_val,
+        final_preds,
+        target_names=["Human", "AI"]
+    )
+)
+
+best_model.save_model("classifier/xgboost_final_model.json")
+joblib.dump(best_model, "classifier/xgboost_final_model.pkl")
+
+print("\n Tuned XGBoost model saved successfully")

featureextraction/feature_concatenation.py

@@ -0,0 +1,56 @@
+import pandas as pd
+import numpy as np
+import os
+
+PROJECT_ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+
+STAT_PATH = os.path.join(PROJECT_ROOT, "featureextraction", "step1_statistical_extraction")
+AST_PATH = os.path.join(PROJECT_ROOT, "featureextraction", "step2_ast_extraction")
+STYLE_PATH = os.path.join(PROJECT_ROOT, "featureextraction", "step3_stylometry_extraction")
+SEM_PATH = os.path.join(PROJECT_ROOT, "featureextraction", "semantic_features")
+
+OUTPUT_DIR = os.path.join(PROJECT_ROOT, "featureextraction", "final_features")
+os.makedirs(OUTPUT_DIR, exist_ok=True)
+
+def load_and_concatenate(split):
+    stat_df = pd.read_csv(os.path.join(STAT_PATH, f"{split}_features.csv"))
+    ast_df = pd.read_csv(os.path.join(AST_PATH, f"{split}_features.csv"))
+    style_df = pd.read_csv(os.path.join(STYLE_PATH, f"{split}_features.csv"))
+
+    y = stat_df["Label"].values
+
+    if "language" in stat_df.columns:
+        lang_onehot = pd.get_dummies(stat_df["language"], prefix="lang")
+    else:
+        raise ValueError("Language column not found for one-hot encoding!")
+
+    X_stat = stat_df.drop(columns=["Label", "language"]).values
+
+    X_ast = ast_df.drop(columns=["Label"]).values
+    X_style = style_df.drop(columns=["Label"]).values
+
+    X_lang = lang_onehot.values
+
+    X_sem = np.load(os.path.join(SEM_PATH, f"{split}_unixcoder.npy"))
+
+    assert (
+        len(X_stat) == len(X_ast) == len(X_style) == len(X_lang) == len(X_sem)
+    ), f"Row mismatch in {split} split!"
+
+  
+    X_final = np.hstack([X_stat, X_ast, X_style, X_lang, X_sem])
+
+    return X_final.astype(float), y
+
+if __name__ == "__main__":
+
+    for split in ["train", "val", "test"]:
+        X, y = load_and_concatenate(split)
+
+        np.save(os.path.join(OUTPUT_DIR, f"{split}_X.npy"), X)
+        np.save(os.path.join(OUTPUT_DIR, f"{split}_y.npy"), y)
+
+        print(f"{split.upper()} FEATURES SHAPE: {X.shape}")
+        print(f"{split.upper()} FEATURE TYPE:", X.dtype)
+
+    print("\nFeature concatenation with ONE-HOT language encoding completed")

featureextraction/semantic_features/unixcoder_embedding.py

@@ -0,0 +1,35 @@
+import torch
+from transformers import AutoTokenizer, AutoModel
+
+MODEL_NAME = "microsoft/unixcoder-base"
+
+tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
+unix_model = AutoModel.from_pretrained(MODEL_NAME)
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+unix_model.to(device)
+unix_model.eval()
+
+
+def get_unixcoder_embedding(code, max_length=512):
+
+    inputs = tokenizer(
+        code,
+        padding=True,
+        truncation=True,
+        max_length=max_length,
+        return_tensors="pt"
+    )
+
+    inputs = {k: v.to(device) for k, v in inputs.items()}
+
+    with torch.no_grad():
+        outputs = unix_model(**inputs)
+        last_hidden = outputs.last_hidden_state
+
+        cls_embedding = last_hidden[:, 0, :]
+        mean_embedding = last_hidden.mean(dim=1)
+
+        combined = torch.cat((cls_embedding, mean_embedding), dim=1)
+
+    return combined.cpu().numpy().flatten()

featureextraction/step1_statistical_extraction/step1_statistical_extraction.py

@@ -0,0 +1,97 @@
+import pandas as pd
+import numpy as np
+import re
+import os
+from collections import Counter
+from math import log2
+
+def tokenize(code):
+    # identifiers, numbers, operators
+    return re.findall(r"[A-Za-z_]+|\d+|==|!=|<=|>=|[+\-*/%]", code)
+
+def token_entropy(tokens):
+    if not tokens:
+        return 0.0
+    counts = Counter(tokens)
+    total = len(tokens)
+    probs = [c / total for c in counts.values()]
+    return -sum(p * log2(p) for p in probs)
+
+def burstiness(tokens):
+    if not tokens:
+        return 0.0
+    counts = Counter(tokens)
+    repeated = sum(c for c in counts.values() if c > 1)
+    return repeated / len(tokens)
+
+def repetition_ratio(tokens):
+    if not tokens:
+        return 0.0
+    return 1 - (len(set(tokens)) / len(tokens))
+
+def avg_token_length(tokens):
+    if not tokens:
+        return 0.0
+    return np.mean([len(t) for t in tokens])
+
+def vocab_richness(tokens):
+    if not tokens:
+        return 0.0
+    return len(set(tokens)) / len(tokens)
+
+def num_tokens(tokens):
+    return len(tokens)
+
+def unique_token_ratio(tokens):
+    if not tokens:
+        return 0.0
+    return len(set(tokens)) / len(tokens)
+
+def avg_line_length(code):
+    lines = code.split("\n")
+    if not lines:
+        return 0.0
+    return np.mean([len(l) for l in lines if l.strip()])
+
+def extract_features(df):
+    features = []
+
+    for _, row in df.iterrows():
+        code = str(row["normalized_code"])
+        tokens = tokenize(code)
+
+        features.append({
+            "entropy": token_entropy(tokens),
+            "burstiness": burstiness(tokens),
+            "repetition_ratio": repetition_ratio(tokens),
+            "avg_token_length": avg_token_length(tokens),
+            "vocab_richness": vocab_richness(tokens),
+
+           
+            "num_tokens": num_tokens(tokens),
+            "unique_token_ratio": unique_token_ratio(tokens),
+            "avg_line_length": avg_line_length(code),
+
+            "language": row.get("Language", "unknown")
+        })
+
+    return pd.DataFrame(features)
+
+if __name__ == "__main__":
+
+    os.makedirs("basemodel", exist_ok=True)
+
+    for split in ["train", "val", "test"]:
+        input_path = f"dataset/processed/dataset_{split}.csv"
+        df = pd.read_csv(input_path)
+
+        if "Label (0- HUMAN, 1-AI)" in df.columns:
+            df = df.rename(columns={"Label (0- HUMAN, 1-AI)": "Label"})
+
+        X = extract_features(df)
+        X["Label"] = df["Label"]
+
+        output_path = f"featureextraction/step1_stastical_extraction/{split}_features.csv"
+        X.to_csv(output_path, index=False)
+
+        print(f"Statistical features extracted for {split}")

featureextraction/step1_statistical_extraction/test_features.csv

@@ -0,0 +1,75 @@
+entropy,burstiness,repetition_ratio,avg_token_length,vocab_richness,num_tokens,unique_token_ratio,avg_line_length,language,Label
+4.83141993487849,0.781021897810219,0.583941605839416,3.153284671532847,0.41605839416058393,137,0.41605839416058393,30.09090909090909,python,1.0
+4.100105160750668,0.8333333333333334,0.6111111111111112,2.7962962962962963,0.3888888888888889,54,0.3888888888888889,33.5,python,0.0
+3.549523459597832,0.7857142857142857,0.5,6.5,0.5,28,0.5,27.818181818181817,java,1.0
+4.598877796136016,0.7543859649122807,0.49122807017543857,3.473684210526316,0.5087719298245614,57,0.5087719298245614,43.666666666666664,python,1.0
+4.008947073809811,0.7708333333333334,0.5625,2.5208333333333335,0.4375,48,0.4375,24.181818181818183,java,1.0
+3.9620439376607153,0.8448275862068966,0.6379310344827587,2.689655172413793,0.3620689655172414,58,0.3620689655172414,24.5,java,0.0
+4.496439344671016,0.65,0.375,4.325,0.625,40,0.625,41.0,python,1.0
+5.038474870244337,0.7261904761904762,0.5119047619047619,4.083333333333333,0.4880952380952381,84,0.4880952380952381,40.916666666666664,python,0.0
+3.828442584567261,0.9724770642201835,0.8256880733944953,2.8990825688073394,0.1743119266055046,109,0.1743119266055046,34.111111111111114,python,1.0
+5.158447453320608,0.7702702702702703,0.45945945945945943,4.581081081081081,0.5405405405405406,74,0.5405405405405406,19.52173913043478,java,0.0
+4.074700960493769,0.6470588235294118,0.5098039215686274,4.823529411764706,0.49019607843137253,51,0.49019607843137253,31.272727272727273,python,0.0
+5.216551170557516,0.32653061224489793,0.18367346938775508,6.489795918367347,0.8163265306122449,49,0.8163265306122449,31.466666666666665,java,1.0
+4.662553238462382,0.8235294117647058,0.5588235294117647,2.926470588235294,0.4411764705882353,68,0.4411764705882353,25.866666666666667,python,1.0
+4.718863069670588,0.8344827586206897,0.6827586206896552,4.703448275862069,0.31724137931034485,145,0.31724137931034485,28.0,java,0.0
+4.186133469144905,0.7878787878787878,0.6060606060606061,3.015151515151515,0.3939393939393939,66,0.3939393939393939,30.636363636363637,python,1.0
+4.201497374489525,0.9393939393939394,0.7474747474747474,1.9292929292929293,0.25252525252525254,99,0.25252525252525254,46.142857142857146,java,1.0
+5.555743206064364,0.7756410256410257,0.5769230769230769,4.397435897435898,0.4230769230769231,156,0.4230769230769231,33.875,java,0.0
+5.120767187212948,0.819672131147541,0.5819672131147541,4.122950819672131,0.4180327868852459,122,0.4180327868852459,27.517241379310345,java,0.0
+5.411695856960623,0.5733333333333334,0.3466666666666667,4.773333333333333,0.6533333333333333,75,0.6533333333333333,44.36363636363637,python,0.0
+5.9364666815595095,0.9323308270676691,0.7969924812030076,4.402255639097745,0.20300751879699247,532,0.20300751879699247,46.348314606741575,java,0.0
+5.921411473217016,0.8119266055045872,0.6146788990825688,5.041284403669724,0.3853211009174312,218,0.3853211009174312,43.567567567567565,python,0.0
+5.681975847739398,0.5217391304347826,0.3369565217391305,6.336956521739131,0.6630434782608695,92,0.6630434782608695,41.10526315789474,python,1.0
+4.9007043302212505,0.7849462365591398,0.5591397849462365,5.236559139784946,0.44086021505376344,93,0.44086021505376344,38.111111111111114,python,0.0
+5.3228285773092585,0.9466019417475728,0.7524271844660194,3.5194174757281553,0.24757281553398058,206,0.24757281553398058,33.48837209302326,java,0.0
+4.886224508551541,0.6811594202898551,0.4492753623188406,4.956521739130435,0.5507246376811594,69,0.5507246376811594,27.904761904761905,java,0.0
+3.7468422480983827,0.7317073170731707,0.5609756097560976,2.5853658536585367,0.43902439024390244,41,0.43902439024390244,23.1,python,1.0
+5.711349874129563,0.9204545454545454,0.7727272727272727,3.7017045454545454,0.22727272727272727,352,0.22727272727272727,33.15068493150685,java,0.0
+4.708644255945031,0.8160919540229885,0.6091954022988506,4.551724137931035,0.39080459770114945,87,0.39080459770114945,38.73684210526316,java,0.0
+4.366729296672175,0.46875,0.28125,6.21875,0.71875,32,0.71875,24.0,java,1.0
+4.819680984926536,0.7704918032786885,0.47540983606557374,3.19672131147541,0.5245901639344263,61,0.5245901639344263,33.18181818181818,python,0.0
+4.982067485400015,0.8495575221238938,0.6283185840707964,2.8849557522123894,0.37168141592920356,113,0.37168141592920356,37.5,python,0.0
+5.792349655949547,0.8646288209606987,0.6506550218340612,6.423580786026201,0.34934497816593885,229,0.34934497816593885,31.1,java,1.0
+5.172672355162728,0.8211382113821138,0.6016260162601625,3.2195121951219514,0.3983739837398374,123,0.3983739837398374,29.7,java,0.0
+4.509883902076781,0.5869565217391305,0.3913043478260869,3.9782608695652173,0.6086956521739131,46,0.6086956521739131,31.6,java,1.0
+5.0781577563714695,0.7582417582417582,0.5274725274725275,4.912087912087912,0.4725274725274725,91,0.4725274725274725,42.94736842105263,java,0.0
+4.752102792050065,0.8555555555555555,0.6222222222222222,2.6,0.37777777777777777,90,0.37777777777777777,30.875,python,1.0
+5.385167845762692,0.6842105263157895,0.4631578947368421,5.442105263157894,0.5368421052631579,95,0.5368421052631579,39.34782608695652,java,0.0
+4.698453748752097,0.8333333333333334,0.6111111111111112,2.3,0.3888888888888889,90,0.3888888888888889,30.75,python,1.0
+5.0936606896881855,0.48,0.26,4.5,0.74,50,0.74,36.44444444444444,python,1.0
+4.762058188070034,0.5689655172413793,0.3793103448275862,3.8793103448275863,0.6206896551724138,58,0.6206896551724138,35.875,python,0.0
+3.842216900031167,0.8703703703703703,0.6666666666666667,3.4444444444444446,0.3333333333333333,54,0.3333333333333333,26.571428571428573,python,1.0
+5.394236721618798,0.8340425531914893,0.6851063829787234,3.1659574468085108,0.3148936170212766,235,0.3148936170212766,45.16129032258065,java,0.0
+3.6368421881310113,0.5416666666666666,0.375,5.916666666666667,0.625,24,0.625,29.857142857142858,python,1.0
+4.371179616205941,0.821917808219178,0.6164383561643836,4.575342465753424,0.3835616438356164,73,0.3835616438356164,31.31578947368421,python,0.0
+5.292298716803247,0.6103896103896104,0.38961038961038963,4.597402597402597,0.6103896103896104,77,0.6103896103896104,43.63636363636363,python,0.0
+4.9019636334438,0.8170731707317073,0.5609756097560976,4.7926829268292686,0.43902439024390244,82,0.43902439024390244,34.21739130434783,java,0.0
+6.249267385727016,0.8602150537634409,0.6774193548387097,5.467741935483871,0.3225806451612903,372,0.3225806451612903,44.231884057971016,python,0.0
+5.022946255731872,0.905982905982906,0.6495726495726496,2.769230769230769,0.3504273504273504,117,0.3504273504273504,34.44,python,0.0
+5.183364648336087,0.578125,0.34375,6.140625,0.65625,64,0.65625,36.5,python,1.0
+3.8514475312208925,0.891566265060241,0.7469879518072289,2.5542168674698793,0.25301204819277107,83,0.25301204819277107,33.15384615384615,java,1.0
+4.76725052491479,0.8088235294117647,0.5294117647058824,4.382352941176471,0.47058823529411764,68,0.47058823529411764,29.857142857142858,java,0.0
+4.053149247727424,0.93,0.76,2.96,0.24,100,0.24,30.523809523809526,java,1.0
+5.508457113311519,0.8251748251748252,0.5944055944055944,5.1328671328671325,0.40559440559440557,143,0.40559440559440557,42.411764705882355,java,0.0
+4.041108871665959,0.7560975609756098,0.5121951219512195,5.048780487804878,0.4878048780487805,41,0.4878048780487805,26.75,java,1.0
+5.264706383768677,0.7769784172661871,0.5827338129496402,4.683453237410072,0.4172661870503597,139,0.4172661870503597,71.57142857142857,python,0.0
+4.888197350707755,0.9251700680272109,0.7210884353741497,2.5374149659863945,0.2789115646258503,147,0.2789115646258503,37.392857142857146,python,0.0
+4.246439344671016,0.65,0.42500000000000004,4.5,0.575,40,0.575,25.818181818181817,python,1.0
+5.3571067941182635,0.4838709677419355,0.27419354838709675,7.532258064516129,0.7258064516129032,62,0.7258064516129032,46.15384615384615,java,1.0
+4.882260151910624,0.8173076923076923,0.625,3.951923076923077,0.375,104,0.375,29.307692307692307,java,0.0
+4.413909765557392,0.40625,0.25,3.84375,0.75,32,0.75,28.666666666666668,python,0.0
+4.003044786734202,0.6666666666666666,0.4242424242424242,4.424242424242424,0.5757575757575758,33,0.5757575757575758,28.714285714285715,python,0.0
+4.138502179294739,0.8235294117647058,0.6176470588235294,3.8088235294117645,0.38235294117647056,68,0.38235294117647056,38.45454545454545,java,1.0
+3.490664945449586,0.8979591836734694,0.6938775510204082,4.571428571428571,0.30612244897959184,49,0.30612244897959184,26.526315789473685,java,0.0
+4.742155477497019,0.7049180327868853,0.47540983606557374,2.80327868852459,0.5245901639344263,61,0.5245901639344263,27.4,python,0.0
+4.907962016217061,0.4318181818181818,0.25,3.0454545454545454,0.75,44,0.75,62.0,python,1.0
+5.299012661837314,0.7024793388429752,0.5371900826446281,3.677685950413223,0.4628099173553719,121,0.4628099173553719,48.0,python,0.0
+5.643744827511846,0.8793774319066148,0.7081712062256809,7.603112840466926,0.2918287937743191,257,0.2918287937743191,36.523809523809526,java,1.0
+3.640223928941852,0.26666666666666666,0.1333333333333333,4.4,0.8666666666666667,15,0.8666666666666667,32.333333333333336,python,1.0
+4.220518125914522,0.9583333333333334,0.8194444444444444,2.0277777777777777,0.18055555555555555,144,0.18055555555555555,41.92307692307692,java,1.0
+5.706414205462762,0.7961783439490446,0.5796178343949044,7.503184713375796,0.42038216560509556,157,0.42038216560509556,38.61363636363637,java,1.0
+5.222228298430333,0.8230088495575221,0.5663716814159292,3.982300884955752,0.4336283185840708,113,0.4336283185840708,41.75,python,0.0
+4.277261292670126,0.8588235294117647,0.6705882352941177,3.0,0.32941176470588235,85,0.32941176470588235,34.35294117647059,python,0.0
+3.8035088547976783,0.5416666666666666,0.33333333333333337,4.208333333333333,0.6666666666666666,24,0.6666666666666666,24.571428571428573,python,1.0
+5.251158101143722,0.8026315789473685,0.618421052631579,4.894736842105263,0.3815789473684211,152,0.3815789473684211,44.06896551724138,java,0.0

featureextraction/step1_statistical_extraction/train_features.csv

@@ -0,0 +1,345 @@
+entropy,burstiness,repetition_ratio,avg_token_length,vocab_richness,num_tokens,unique_token_ratio,avg_line_length,language,Label
+5.207956079154271,0.8712121212121212,0.6515151515151515,3.9015151515151514,0.3484848484848485,132,0.3484848484848485,41.464285714285715,python,0.0
+4.844227218082805,0.8962962962962963,0.7037037037037037,3.1259259259259258,0.2962962962962963,135,0.2962962962962963,15.704545454545455,java,0.0
+5.110041502517433,0.916083916083916,0.6783216783216783,4.5174825174825175,0.32167832167832167,143,0.32167832167832167,25.784313725490197,java,0.0
+4.599986090068817,0.8839285714285714,0.7053571428571428,4.616071428571429,0.29464285714285715,112,0.29464285714285715,30.68421052631579,java,0.0
+4.469392725378344,0.8589743589743589,0.6153846153846154,3.230769230769231,0.38461538461538464,78,0.38461538461538464,23.923076923076923,java,0.0
+4.293744383186907,0.859375,0.625,2.953125,0.375,64,0.375,26.210526315789473,java,0.0
+4.549676205034737,0.860655737704918,0.6885245901639344,5.918032786885246,0.3114754098360656,122,0.3114754098360656,33.96666666666667,python,0.0
+4.328596578632476,0.8163265306122449,0.5306122448979591,3.2244897959183674,0.46938775510204084,49,0.46938775510204084,28.333333333333332,python,0.0
+4.597753650865813,0.7941176470588235,0.5588235294117647,3.0441176470588234,0.4411764705882353,68,0.4411764705882353,30.916666666666668,java,1.0
+4.0675361498812554,0.8333333333333334,0.6666666666666667,2.0384615384615383,0.3333333333333333,78,0.3333333333333333,28.0,python,0.0
+4.853956434293166,0.898989898989899,0.6464646464646464,4.101010101010101,0.35353535353535354,99,0.35353535353535354,31.40740740740741,java,0.0
+6.10213285294983,0.875,0.6736111111111112,7.423611111111111,0.3263888888888889,288,0.3263888888888889,37.31395348837209,java,1.0
+3.053791405126089,0.7777777777777778,0.6388888888888888,4.027777777777778,0.3611111111111111,36,0.3611111111111111,22.25,python,1.0
+4.379556754405944,0.8620689655172413,0.5689655172413793,4.120689655172414,0.43103448275862066,58,0.43103448275862066,35.36363636363637,java,1.0
+4.538844415918181,0.8026315789473685,0.5921052631578947,3.3026315789473686,0.40789473684210525,76,0.40789473684210525,15.043478260869565,java,0.0
+4.435455286899907,0.847457627118644,0.576271186440678,2.694915254237288,0.423728813559322,59,0.423728813559322,26.818181818181817,python,1.0
+4.1277054819279435,0.8333333333333334,0.5625,3.0,0.4375,48,0.4375,24.6,java,1.0
+4.498419401169,0.7058823529411765,0.47058823529411764,3.2941176470588234,0.5294117647058824,51,0.5294117647058824,25.866666666666667,java,0.0
+3.615922063835167,0.7333333333333333,0.5,5.433333333333334,0.5,30,0.5,30.125,python,0.0
+5.043983303761563,0.5833333333333334,0.3666666666666667,4.983333333333333,0.6333333333333333,60,0.6333333333333333,37.5,python,1.0
+3.6897037321995474,0.42857142857142855,0.2857142857142857,7.095238095238095,0.7142857142857143,21,0.7142857142857143,22.7,java,1.0
+3.791950834682379,0.7142857142857143,0.5934065934065934,2.769230769230769,0.4065934065934066,91,0.4065934065934066,34.0,python,1.0
+5.542412227395111,0.7934782608695652,0.6141304347826086,3.6793478260869565,0.3858695652173913,184,0.3858695652173913,61.75,java,0.0
+3.8463189626846366,0.8947368421052632,0.5526315789473684,3.0526315789473686,0.4473684210526316,38,0.4473684210526316,20.6,python,1.0
+4.7720552088742005,0.6666666666666666,0.375,6.229166666666667,0.625,48,0.625,29.214285714285715,java,1.0
+5.085703343051975,0.7099236641221374,0.549618320610687,3.931297709923664,0.45038167938931295,131,0.45038167938931295,32.40909090909091,python,0.0
+4.253863368422077,0.6578947368421053,0.42105263157894735,3.6052631578947367,0.5789473684210527,38,0.5789473684210527,34.833333333333336,java,1.0
+5.285942953673669,0.8157894736842105,0.5614035087719298,3.4035087719298245,0.43859649122807015,114,0.43859649122807015,45.666666666666664,java,0.0
+4.324534762707879,0.7111111111111111,0.4666666666666667,8.177777777777777,0.5333333333333333,45,0.5333333333333333,37.30769230769231,java,1.0
+4.895714600514266,0.8457943925233645,0.7383177570093458,2.6448598130841123,0.2616822429906542,214,0.2616822429906542,18.784313725490197,java,0.0
+3.327339900083509,0.6666666666666666,0.4285714285714286,3.0952380952380953,0.5714285714285714,21,0.5714285714285714,22.8,python,1.0
+5.581633377149314,0.5934065934065934,0.37362637362637363,6.318681318681318,0.6263736263736264,91,0.6263736263736264,40.78947368421053,python,1.0
+5.657995400316549,0.8997289972899729,0.7425474254742548,3.907859078590786,0.25745257452574527,369,0.25745257452574527,56.26315789473684,java,0.0
+4.029187712039474,0.8367346938775511,0.5918367346938775,2.9183673469387754,0.40816326530612246,49,0.40816326530612246,30.0,java,1.0
+4.8582227682726,0.9393939393939394,0.7348484848484849,2.6363636363636362,0.26515151515151514,132,0.26515151515151514,32.0,python,1.0
+4.42309119581308,0.7941176470588235,0.5735294117647058,2.5588235294117645,0.4264705882352941,68,0.4264705882352941,30.454545454545453,java,1.0
+3.7665853954588724,0.7704918032786885,0.639344262295082,2.8852459016393444,0.36065573770491804,61,0.36065573770491804,33.90909090909091,python,0.0
+5.197908393371423,0.8072289156626506,0.6325301204819277,5.0843373493975905,0.3674698795180723,166,0.3674698795180723,38.65625,java,0.0
+4.536286231168867,0.42857142857142855,0.2571428571428571,3.6,0.7428571428571429,35,0.7428571428571429,31.166666666666668,python,1.0
+4.709586788340617,0.8957055214723927,0.7300613496932515,3.361963190184049,0.26993865030674846,163,0.26993865030674846,17.816326530612244,java,0.0
+4.5049847535671015,0.9545454545454546,0.7897727272727273,3.3806818181818183,0.21022727272727273,176,0.21022727272727273,36.16129032258065,java,0.0
+4.9423908406293835,0.6538461538461539,0.4358974358974359,3.8333333333333335,0.5641025641025641,78,0.5641025641025641,45.125,python,0.0
+4.41120441892766,0.7777777777777778,0.4666666666666667,6.488888888888889,0.5333333333333333,45,0.5333333333333333,44.77777777777778,python,0.0
+5.262208612706845,0.7327586206896551,0.5431034482758621,4.146551724137931,0.45689655172413796,116,0.45689655172413796,28.666666666666668,python,1.0
+4.208410187268525,0.48148148148148145,0.2592592592592593,4.222222222222222,0.7407407407407407,27,0.7407407407407407,32.2,python,1.0
+5.793376775298043,0.8823529411764706,0.680672268907563,5.529411764705882,0.31932773109243695,238,0.31932773109243695,27.633333333333333,java,1.0
+5.691854216621956,0.7849462365591398,0.5967741935483871,4.973118279569892,0.4032258064516129,186,0.4032258064516129,40.333333333333336,python,0.0
+5.058984089445427,0.3333333333333333,0.16666666666666663,6.785714285714286,0.8333333333333334,42,0.8333333333333334,30.923076923076923,python,1.0
+4.433157434964864,0.9736842105263158,0.7543859649122807,2.4035087719298245,0.24561403508771928,114,0.24561403508771928,31.5,python,1.0
+5.268123017576671,0.6164383561643836,0.3835616438356164,4.917808219178082,0.6164383561643836,73,0.6164383561643836,44.09090909090909,python,0.0
+6.198108865484992,0.7208121827411168,0.5025380710659899,5.0456852791878175,0.49746192893401014,197,0.49746192893401014,41.638888888888886,python,1.0
+4.5615159206608356,0.8227848101265823,0.6075949367088608,3.5569620253164556,0.3924050632911392,79,0.3924050632911392,13.678571428571429,java,0.0
+5.04371468310881,0.6351351351351351,0.43243243243243246,4.1891891891891895,0.5675675675675675,74,0.5675675675675675,37.0,java,0.0
+3.995906598484245,0.6944444444444444,0.4722222222222222,4.333333333333333,0.5277777777777778,36,0.5277777777777778,27.5,python,1.0
+4.757950193468896,0.9148936170212766,0.648936170212766,2.893617021276596,0.35106382978723405,94,0.35106382978723405,34.0,python,1.0
+5.053560274699695,0.9136690647482014,0.7050359712230216,4.705035971223022,0.2949640287769784,139,0.2949640287769784,39.69565217391305,python,0.0
+3.833323296470577,0.8813559322033898,0.6779661016949152,2.288135593220339,0.3220338983050847,59,0.3220338983050847,27.5,java,1.0
+4.253212018409155,0.5882352941176471,0.3529411764705882,4.117647058823529,0.6470588235294118,34,0.6470588235294118,13.75,java,0.0
+4.611949334080443,0.5384615384615384,0.3076923076923077,4.717948717948718,0.6923076923076923,39,0.6923076923076923,34.42857142857143,python,0.0
+4.756182308560408,0.9186991869918699,0.7073170731707317,3.3089430894308944,0.2926829268292683,123,0.2926829268292683,28.09090909090909,java,0.0
+5.1256959966316895,0.9022988505747126,0.6954022988505747,3.8045977011494254,0.3045977011494253,174,0.3045977011494253,31.727272727272727,java,0.0
+4.274386660093705,0.8103448275862069,0.5862068965517242,2.2413793103448274,0.41379310344827586,58,0.41379310344827586,39.166666666666664,java,1.0
+5.649232493101554,0.8258064516129032,0.5870967741935484,5.961290322580645,0.4129032258064516,155,0.4129032258064516,29.104166666666668,java,0.0
+3.8005595576361078,0.875,0.6785714285714286,3.0892857142857144,0.32142857142857145,56,0.32142857142857145,24.357142857142858,python,1.0
+4.635290969626768,0.6666666666666666,0.43137254901960786,3.7254901960784315,0.5686274509803921,51,0.5686274509803921,25.46153846153846,java,0.0
+4.854346937661789,0.5740740740740741,0.37037037037037035,5.0,0.6296296296296297,54,0.6296296296296297,25.4,java,0.0
+4.393061650825727,0.5,0.2941176470588235,7.352941176470588,0.7058823529411765,34,0.7058823529411765,40.625,python,1.0
+4.422442136473173,0.5135135135135135,0.32432432432432434,5.081081081081081,0.6756756756756757,37,0.6756756756756757,39.166666666666664,python,1.0
+4.207435516759526,0.6571428571428571,0.4,3.0285714285714285,0.6,35,0.6,22.375,python,1.0
+5.99203259332159,0.7687074829931972,0.48299319727891155,5.122448979591836,0.5170068027210885,147,0.5170068027210885,39.357142857142854,python,1.0
+6.007074413338004,0.711764705882353,0.5058823529411764,4.129411764705883,0.49411764705882355,170,0.49411764705882355,58.0,python,0.0
+3.6464393446710157,0.5,0.30000000000000004,3.85,0.7,20,0.7,27.0,python,1.0
+4.497874075147441,0.5333333333333333,0.37777777777777777,4.311111111111111,0.6222222222222222,45,0.6222222222222222,30.636363636363637,java,1.0
+4.728255941527934,0.9318181818181818,0.7348484848484849,2.6363636363636362,0.26515151515151514,132,0.26515151515151514,33.68,python,0.0
+4.2354799985372935,0.6862745098039216,0.5098039215686274,2.9411764705882355,0.49019607843137253,51,0.49019607843137253,28.333333333333332,python,0.0
+5.560363930440956,0.8622754491017964,0.6347305389221557,4.029940119760479,0.3652694610778443,167,0.3652694610778443,31.70212765957447,java,0.0
+6.014883786593109,0.5323741007194245,0.38129496402877694,4.863309352517986,0.6187050359712231,139,0.6187050359712231,34.25,python,1.0
+4.384183719779189,0.55,0.375,6.35,0.625,40,0.625,32.916666666666664,java,1.0
+4.36346329043209,0.7105263157894737,0.39473684210526316,6.157894736842105,0.6052631578947368,38,0.6052631578947368,27.733333333333334,java,1.0
+5.92272462852426,0.7397959183673469,0.5510204081632653,4.969387755102041,0.4489795918367347,196,0.4489795918367347,72.89473684210526,java,0.0
+4.17415473143563,0.9078947368421053,0.7105263157894737,2.1578947368421053,0.2894736842105263,76,0.2894736842105263,30.692307692307693,python,1.0
+5.271209951802516,0.9140271493212669,0.7330316742081449,4.97737556561086,0.2669683257918552,221,0.2669683257918552,43.416666666666664,java,0.0
+5.014527492014123,0.9281045751633987,0.7320261437908497,2.7777777777777777,0.2679738562091503,153,0.2679738562091503,38.869565217391305,python,0.0
+4.482483990770269,0.9565217391304348,0.7391304347826086,4.026086956521739,0.2608695652173913,115,0.2608695652173913,36.90909090909091,java,1.0
+5.722786643861381,0.8493975903614458,0.5783132530120482,3.7710843373493974,0.42168674698795183,166,0.42168674698795183,19.595744680851062,java,0.0
+5.300160551498397,0.5909090909090909,0.33333333333333337,5.757575757575758,0.6666666666666666,66,0.6666666666666666,36.75,java,1.0
+4.45148902643068,0.5,0.3157894736842105,3.4210526315789473,0.6842105263157895,38,0.6842105263157895,21.7,python,1.0
+5.072571258365157,0.8296703296703297,0.6428571428571428,3.4450549450549453,0.35714285714285715,182,0.35714285714285715,28.515151515151516,python,1.0
+4.168243795012672,0.5161290322580645,0.32258064516129037,6.0,0.6774193548387096,31,0.6774193548387096,31.375,java,0.0
+4.123939173306749,0.6511627906976745,0.4651162790697675,3.6744186046511627,0.5348837209302325,43,0.5348837209302325,28.2,java,1.0
+4.534110630792036,0.8888888888888888,0.6666666666666667,3.1222222222222222,0.3333333333333333,90,0.3333333333333333,35.3,java,0.0
+4.889260316971673,1.0,0.7348484848484849,3.6515151515151514,0.26515151515151514,132,0.26515151515151514,23.066666666666666,java,0.0
+4.883002384777129,0.9366197183098591,0.7323943661971831,2.852112676056338,0.2676056338028169,142,0.2676056338028169,36.25806451612903,python,0.0
+3.8200705941688406,0.9666666666666667,0.8111111111111111,5.011111111111111,0.18888888888888888,90,0.18888888888888888,37.421052631578945,java,1.0
+5.866907260896749,0.7725321888412017,0.5965665236051503,5.141630901287554,0.4034334763948498,233,0.4034334763948498,71.04347826086956,python,0.0
+4.592226257370053,0.7936507936507936,0.5396825396825398,3.5714285714285716,0.4603174603174603,63,0.4603174603174603,31.5,python,0.0
+4.979264809390598,0.2926829268292683,0.1707317073170732,6.317073170731708,0.8292682926829268,41,0.8292682926829268,33.416666666666664,java,1.0
+3.1693150678026982,0.8705882352941177,0.7294117647058824,2.1411764705882352,0.27058823529411763,85,0.27058823529411763,29.88888888888889,python,1.0
+4.718291773866265,0.8803418803418803,0.6837606837606838,2.871794871794872,0.3162393162393162,117,0.3162393162393162,33.17857142857143,python,0.0
+6.057756669289524,0.5080645161290323,0.3306451612903226,5.588709677419355,0.6693548387096774,124,0.6693548387096774,37.57692307692308,python,1.0
+4.94407911950056,0.8181818181818182,0.5324675324675325,7.077922077922078,0.4675324675324675,77,0.4675324675324675,37.05,java,1.0
+4.726143834398879,0.6136363636363636,0.34090909090909094,4.840909090909091,0.6590909090909091,44,0.6590909090909091,28.25,python,1.0
+4.475969855129791,0.8305084745762712,0.5593220338983051,3.389830508474576,0.4406779661016949,59,0.4406779661016949,27.0,java,1.0
+5.689839588541731,0.783410138248848,0.6082949308755761,3.640552995391705,0.391705069124424,217,0.391705069124424,31.355555555555554,java,0.0
+4.518215514675149,0.9473684210526315,0.6578947368421053,2.526315789473684,0.34210526315789475,76,0.34210526315789475,44.285714285714285,python,1.0
+4.124232718690979,0.7678571428571429,0.5892857142857143,3.1607142857142856,0.4107142857142857,56,0.4107142857142857,25.3125,java,0.0
+4.112426730000285,0.8846153846153846,0.7403846153846154,2.951923076923077,0.25961538461538464,104,0.25961538461538464,34.09090909090909,python,0.0
+5.039480564148708,0.7866666666666666,0.4933333333333333,4.093333333333334,0.5066666666666667,75,0.5066666666666667,32.111111111111114,java,0.0
+4.677139104850461,0.8181818181818182,0.5454545454545454,3.621212121212121,0.45454545454545453,66,0.45454545454545453,22.0,python,0.0
+4.131300342505361,0.7096774193548387,0.3870967741935484,6.161290322580645,0.6129032258064516,31,0.6129032258064516,31.625,java,1.0
+5.462480227696889,0.7959183673469388,0.5850340136054422,4.707482993197279,0.41496598639455784,147,0.41496598639455784,24.872340425531913,java,0.0
+3.6914280318460246,0.8125,0.53125,2.09375,0.46875,32,0.46875,23.333333333333332,python,1.0
+4.214114119970501,0.7142857142857143,0.5102040816326531,2.816326530612245,0.4897959183673469,49,0.4897959183673469,22.583333333333332,java,1.0
+5.697217023615801,0.8171428571428572,0.5828571428571429,5.154285714285714,0.41714285714285715,175,0.41714285714285715,37.93023255813954,java,0.0
+4.0791433740260095,0.6571428571428571,0.4285714285714286,5.057142857142857,0.5714285714285714,35,0.5714285714285714,25.0,python,1.0
+3.7489948035250964,0.47619047619047616,0.2857142857142857,3.4285714285714284,0.7142857142857143,21,0.7142857142857143,26.25,python,1.0
+4.907735246015272,0.75,0.525,4.2125,0.475,80,0.475,29.869565217391305,python,0.0
+4.822037278990095,0.7808219178082192,0.5342465753424658,2.958904109589041,0.4657534246575342,73,0.4657534246575342,33.0,python,0.0
+5.030104271105216,0.8301886792452831,0.6037735849056604,3.349056603773585,0.39622641509433965,106,0.39622641509433965,33.958333333333336,java,1.0
+4.816108939837481,0.7142857142857143,0.4285714285714286,4.392857142857143,0.5714285714285714,56,0.5714285714285714,29.23076923076923,python,1.0
+2.4276042408847402,0.8461538461538461,0.7230769230769231,1.7692307692307692,0.27692307692307694,65,0.27692307692307694,27.666666666666668,python,1.0
+6.584083875285213,0.8148148148148148,0.6271604938271604,6.212345679012346,0.3728395061728395,405,0.3728395061728395,40.2247191011236,python,0.0
+4.1673958618292355,0.8536585365853658,0.6951219512195121,2.7560975609756095,0.3048780487804878,82,0.3048780487804878,29.65,java,0.0
+3.741962789037907,0.6896551724137931,0.4482758620689655,2.0689655172413794,0.5517241379310345,29,0.5517241379310345,18.285714285714285,python,1.0
+4.86754794914527,0.5178571428571429,0.3571428571428571,3.375,0.6428571428571429,56,0.6428571428571429,19.2,java,0.0
+3.9331695057515157,0.9285714285714286,0.5952380952380952,3.5714285714285716,0.40476190476190477,42,0.40476190476190477,38.166666666666664,python,1.0
+4.196483601699262,0.49019607843137253,0.4117647058823529,6.254901960784314,0.5882352941176471,51,0.5882352941176471,34.25,python,0.0
+5.906626836308724,0.8967254408060453,0.7481108312342569,3.44080604534005,0.2518891687657431,397,0.2518891687657431,53.51219512195122,java,0.0
+6.369927477083823,0.5966850828729282,0.4088397790055248,4.900552486187845,0.5911602209944752,181,0.5911602209944752,49.65384615384615,python,1.0
+4.384795487373134,0.5277777777777778,0.33333333333333337,4.611111111111111,0.6666666666666666,36,0.6666666666666666,18.25,java,0.0
+5.139173834043169,0.8348623853211009,0.5871559633027523,4.935779816513762,0.41284403669724773,109,0.41284403669724773,35.370370370370374,java,0.0
+4.238521032323433,0.7708333333333334,0.5208333333333333,4.5625,0.4791666666666667,48,0.4791666666666667,27.571428571428573,python,0.0
+4.460619643646408,0.8309859154929577,0.5915492957746479,4.704225352112676,0.4084507042253521,71,0.4084507042253521,35.72222222222222,java,0.0
+5.2878758225325395,0.9408450704225352,0.8169014084507042,4.366197183098592,0.18309859154929578,355,0.18309859154929578,42.14545454545455,java,0.0
+4.8335434078297626,0.6666666666666666,0.4545454545454546,4.46969696969697,0.5454545454545454,66,0.5454545454545454,65.66666666666667,python,0.0
+3.3889821967054377,0.9555555555555556,0.7333333333333334,5.022222222222222,0.26666666666666666,45,0.26666666666666666,53.0,java,1.0
+5.626267512421823,0.8212290502793296,0.6145251396648045,3.8491620111731844,0.3854748603351955,179,0.3854748603351955,55.166666666666664,python,1.0
+4.5054500825687525,0.9090909090909091,0.696969696969697,3.6464646464646466,0.30303030303030304,99,0.30303030303030304,41.36363636363637,python,0.0
+4.66816999226466,0.5609756097560976,0.31707317073170727,4.7560975609756095,0.6829268292682927,41,0.6829268292682927,30.2,java,1.0
+4.1138276215159735,0.8461538461538461,0.6307692307692307,2.876923076923077,0.36923076923076925,65,0.36923076923076925,31.071428571428573,java,0.0
+3.7489948035250964,0.47619047619047616,0.2857142857142857,6.904761904761905,0.7142857142857143,21,0.7142857142857143,33.57142857142857,java,1.0
+4.639932288803049,0.8472222222222222,0.6944444444444444,3.9166666666666665,0.3055555555555556,144,0.3055555555555556,40.34615384615385,java,0.0
+5.503071993063162,0.7610619469026548,0.504424778761062,3.2920353982300883,0.49557522123893805,113,0.49557522123893805,51.36363636363637,python,1.0
+5.320245758154888,0.8130081300813008,0.5853658536585367,2.227642276422764,0.4146341463414634,123,0.4146341463414634,33.5625,python,1.0
+4.448860987743577,0.5789473684210527,0.3421052631578947,4.131578947368421,0.6578947368421053,38,0.6578947368421053,32.125,python,1.0
+4.526180668235663,0.5897435897435898,0.33333333333333337,4.17948717948718,0.6666666666666666,39,0.6666666666666666,26.9,java,1.0
+4.202589101773195,0.9411764705882353,0.7294117647058824,1.7058823529411764,0.27058823529411763,85,0.27058823529411763,31.3,java,1.0
+5.934435083287192,0.6229508196721312,0.4016393442622951,5.0,0.5983606557377049,122,0.5983606557377049,36.0,python,1.0
+4.713707321865714,0.6885245901639344,0.4590163934426229,6.459016393442623,0.5409836065573771,61,0.5409836065573771,33.8125,python,0.0
+4.3180873032876175,0.9,0.7111111111111111,2.2777777777777777,0.28888888888888886,90,0.28888888888888886,38.833333333333336,python,0.0
+4.285367908896077,0.6875,0.47916666666666663,4.520833333333333,0.5208333333333334,48,0.5208333333333334,40.875,java,1.0
+4.366596139176847,0.6585365853658537,0.41463414634146345,5.609756097560975,0.5853658536585366,41,0.5853658536585366,34.8,java,1.0
+5.343782087298768,0.8652482269503546,0.624113475177305,5.7304964539007095,0.375886524822695,141,0.375886524822695,40.40625,java,0.0
+4.327819531114783,0.375,0.25,6.28125,0.75,32,0.75,29.5,java,1.0
+3.8033524347703986,0.7105263157894737,0.5,2.763157894736842,0.5,38,0.5,26.375,python,0.0
+5.72576174609731,0.797979797979798,0.6161616161616161,3.984848484848485,0.3838383838383838,198,0.3838383838383838,42.55882352941177,java,0.0
+5.409509758627718,0.841025641025641,0.676923076923077,5.671794871794872,0.3230769230769231,195,0.3230769230769231,34.90384615384615,python,0.0
+5.679736241553765,0.7013888888888888,0.5138888888888888,4.743055555555555,0.4861111111111111,144,0.4861111111111111,52.55555555555556,python,0.0
+4.4912054473944405,0.8148148148148148,0.5185185185185186,5.277777777777778,0.48148148148148145,54,0.48148148148148145,36.23076923076923,java,0.0
+3.9362434129830604,0.6333333333333333,0.4,3.1333333333333333,0.6,30,0.6,24.75,python,0.0
+4.215928174708332,0.8852459016393442,0.6229508196721312,2.721311475409836,0.3770491803278688,61,0.3770491803278688,27.923076923076923,python,0.0
+5.255072207410051,0.9081081081081082,0.7135135135135136,3.675675675675676,0.2864864864864865,185,0.2864864864864865,38.34375,python,0.0
+4.8132935840925875,0.8902439024390244,0.5853658536585367,4.365853658536586,0.4146341463414634,82,0.4146341463414634,39.166666666666664,python,0.0
+3.728999521334141,0.9047619047619048,0.7301587301587302,4.619047619047619,0.2698412698412698,63,0.2698412698412698,34.84615384615385,java,1.0
+4.4003105301349255,0.6888888888888889,0.4444444444444444,2.5555555555555554,0.5555555555555556,45,0.5555555555555556,21.90909090909091,python,1.0
+5.192613181399552,0.8343949044585988,0.6624203821656052,3.445859872611465,0.3375796178343949,157,0.3375796178343949,42.476190476190474,python,0.0
+4.896037310447887,0.8761061946902655,0.6548672566371682,3.2123893805309733,0.34513274336283184,113,0.34513274336283184,27.653846153846153,java,0.0
+4.652782517460112,0.6447368421052632,0.513157894736842,2.8684210526315788,0.4868421052631579,76,0.4868421052631579,32.76470588235294,java,0.0
+3.757539644379974,0.9032258064516129,0.7096774193548387,1.6612903225806452,0.2903225806451613,62,0.2903225806451613,24.375,java,1.0
+5.4053549514703,0.7605633802816901,0.5633802816901409,5.640845070422535,0.43661971830985913,142,0.43661971830985913,38.833333333333336,java,0.0
+4.715163766425403,0.9142857142857143,0.6761904761904762,4.580952380952381,0.3238095238095238,105,0.3238095238095238,27.733333333333334,java,0.0
+5.048749755702801,0.8085106382978723,0.5638297872340425,2.74468085106383,0.43617021276595747,94,0.43617021276595747,34.375,java,1.0
+4.412142047769037,0.734375,0.546875,2.9375,0.453125,64,0.453125,31.0,python,0.0
+4.162294909570876,0.4444444444444444,0.2592592592592593,4.148148148148148,0.7407407407407407,27,0.7407407407407407,21.428571428571427,java,0.0
+5.17784699780413,0.46153846153846156,0.32307692307692304,3.4153846153846152,0.676923076923077,65,0.676923076923077,38.92307692307692,java,0.0
+5.4865950410391395,0.8578199052132701,0.6872037914691943,4.1753554502369665,0.3127962085308057,211,0.3127962085308057,37.65217391304348,java,0.0
+5.104500026854035,0.875,0.6333333333333333,4.291666666666667,0.36666666666666664,120,0.36666666666666664,39.80769230769231,java,0.0
+6.146771906965858,0.8115183246073299,0.5445026178010471,5.350785340314136,0.45549738219895286,191,0.45549738219895286,40.8,python,0.0
+4.898714188796334,0.8387096774193549,0.6021505376344086,4.559139784946237,0.3978494623655914,93,0.3978494623655914,40.0,java,0.0
+4.33379987011787,0.6521739130434783,0.4565217391304348,4.195652173913044,0.5434782608695652,46,0.5434782608695652,21.944444444444443,java,0.0
+4.720077865655612,0.8817204301075269,0.6344086021505376,4.806451612903226,0.3655913978494624,93,0.3655913978494624,35.95238095238095,java,0.0
+4.971145882958361,0.543859649122807,0.3508771929824561,4.0701754385964914,0.6491228070175439,57,0.6491228070175439,40.375,python,1.0
+4.608465630737972,0.89,0.6799999999999999,2.32,0.32,100,0.32,30.428571428571427,java,1.0
+4.553348286183388,0.8556701030927835,0.6597938144329897,5.391752577319588,0.3402061855670103,97,0.3402061855670103,31.903225806451612,java,0.0
+5.111219097255348,0.6422018348623854,0.5045871559633027,4.009174311926605,0.4954128440366973,109,0.4954128440366973,29.88888888888889,java,0.0
+5.169436190447377,0.8862275449101796,0.7005988023952097,2.8862275449101795,0.2994011976047904,167,0.2994011976047904,42.2,java,1.0
+4.897236854659848,0.6271186440677966,0.4067796610169492,4.593220338983051,0.5932203389830508,59,0.5932203389830508,29.384615384615383,java,0.0
+5.5494840429514545,0.9004739336492891,0.6966824644549763,6.156398104265403,0.3033175355450237,211,0.3033175355450237,33.016949152542374,java,0.0
+4.316499122181017,0.8412698412698413,0.6031746031746033,2.6507936507936507,0.3968253968253968,63,0.3968253968253968,29.818181818181817,java,1.0
+4.946526678933779,0.5,0.31481481481481477,4.203703703703703,0.6851851851851852,54,0.6851851851851852,48.5,python,1.0
+3.980473809580999,0.9195402298850575,0.7471264367816092,2.528735632183908,0.25287356321839083,87,0.25287356321839083,22.53846153846154,java,0.0
+4.931343145380855,0.8930817610062893,0.6666666666666667,2.0440251572327046,0.3333333333333333,159,0.3333333333333333,25.423076923076923,python,1.0
+5.865359596361463,0.7432432432432432,0.5067567567567568,5.256756756756757,0.49324324324324326,148,0.49324324324324326,34.970588235294116,python,1.0
+5.124006295628484,0.45614035087719296,0.2807017543859649,3.754385964912281,0.7192982456140351,57,0.7192982456140351,38.0,python,0.0
+6.085910006150067,0.8445378151260504,0.6218487394957983,6.142857142857143,0.37815126050420167,238,0.37815126050420167,28.604395604395606,java,1.0
+4.868510971586463,0.7411764705882353,0.5411764705882354,3.2470588235294118,0.4588235294117647,85,0.4588235294117647,28.17391304347826,java,0.0
+4.605408294302115,0.8611111111111112,0.5972222222222222,3.138888888888889,0.4027777777777778,72,0.4027777777777778,33.857142857142854,python,0.0
+4.73995425222158,0.6964285714285714,0.4464285714285714,6.482142857142857,0.5535714285714286,56,0.5535714285714286,33.4375,java,1.0
+4.8675567139411235,0.7701149425287356,0.5517241379310345,4.816091954022989,0.4482758620689655,87,0.4482758620689655,36.13636363636363,java,0.0
+4.246481972040831,0.7454545454545455,0.5272727272727273,4.2727272727272725,0.4727272727272727,55,0.4727272727272727,30.416666666666668,python,0.0
+2.8016094970590277,0.8,0.55,5.0,0.45,20,0.45,28.8,python,1.0
+5.7659039797576215,0.5669291338582677,0.40944881889763785,5.377952755905512,0.5905511811023622,127,0.5905511811023622,35.2962962962963,java,0.0
+4.8369927564722754,0.5978260869565217,0.46739130434782605,5.695652173913044,0.532608695652174,92,0.532608695652174,30.0,python,1.0
+5.650410881834686,0.8103448275862069,0.6091954022988506,4.022988505747127,0.39080459770114945,174,0.39080459770114945,39.65853658536585,java,0.0
+4.33611521033683,0.6363636363636364,0.43181818181818177,4.75,0.5681818181818182,44,0.5681818181818182,34.6,java,1.0
+6.041274356789173,0.8076923076923077,0.657051282051282,3.801282051282051,0.34294871794871795,312,0.34294871794871795,30.096385542168676,java,0.0
+5.289310575466358,0.6428571428571429,0.5,4.190476190476191,0.5,126,0.5,31.615384615384617,python,1.0
+4.700186536808578,0.7741935483870968,0.5,3.532258064516129,0.5,62,0.5,33.76923076923077,python,0.0
+4.180451390892101,0.4444444444444444,0.2592592592592593,4.777777777777778,0.7407407407407407,27,0.7407407407407407,22.25,python,1.0
+4.786213536530154,0.9016393442622951,0.6967213114754098,2.762295081967213,0.30327868852459017,122,0.30327868852459017,29.083333333333332,python,1.0
+5.081486951808688,0.7666666666666667,0.5222222222222221,3.7888888888888888,0.4777777777777778,90,0.4777777777777778,35.8,python,1.0
+4.44849995084568,0.8461538461538461,0.6593406593406593,3.5934065934065935,0.34065934065934067,91,0.34065934065934067,31.045454545454547,python,0.0
+4.7077148025974385,0.37142857142857144,0.19999999999999996,7.228571428571429,0.8,35,0.8,40.888888888888886,java,1.0
+4.51662508214898,0.8837209302325582,0.6627906976744187,3.302325581395349,0.3372093023255814,86,0.3372093023255814,23.25,java,0.0
+4.591236738386586,0.7903225806451613,0.532258064516129,4.193548387096774,0.46774193548387094,62,0.46774193548387094,26.352941176470587,java,1.0
+4.770332655554241,0.7123287671232876,0.5205479452054795,2.5753424657534247,0.4794520547945205,73,0.4794520547945205,14.681818181818182,java,0.0
+4.193064830243576,0.9555555555555556,0.7333333333333334,2.3555555555555556,0.26666666666666666,90,0.26666666666666666,24.68421052631579,python,1.0
+3.935732282211645,0.90625,0.703125,3.1875,0.296875,64,0.296875,34.72727272727273,java,1.0
+5.255459047825728,0.7096774193548387,0.4838709677419355,4.32258064516129,0.5161290322580645,93,0.5161290322580645,47.583333333333336,python,0.0
+5.56295430052689,0.9239543726235742,0.7490494296577948,6.391634980988593,0.2509505703422053,263,0.2509505703422053,51.39655172413793,java,0.0
+4.7624023788690275,0.8846153846153846,0.5897435897435898,3.41025641025641,0.41025641025641024,78,0.41025641025641024,27.95,python,0.0
+4.641341251556263,0.7164179104477612,0.5223880597014925,3.5671641791044775,0.47761194029850745,67,0.47761194029850745,32.69230769230769,java,1.0
+3.9580339180966533,0.9142857142857143,0.7142857142857143,4.014285714285714,0.2857142857142857,70,0.2857142857142857,24.5,python,1.0
+3.821296333684861,0.6071428571428571,0.3928571428571429,4.357142857142857,0.6071428571428571,28,0.6071428571428571,26.5,java,1.0
+5.99080968202175,0.8926553672316384,0.7033898305084746,4.759887005649717,0.2966101694915254,354,0.2966101694915254,40.125,python,0.0
+5.023079206577155,0.8592592592592593,0.674074074074074,2.77037037037037,0.32592592592592595,135,0.32592592592592595,31.88888888888889,java,0.0
+4.427567157116929,0.675,0.4,5.55,0.6,40,0.6,31.4,java,1.0
+4.183636205660773,0.8666666666666667,0.6916666666666667,2.8583333333333334,0.30833333333333335,120,0.30833333333333335,28.38888888888889,python,1.0
+4.6597716630131325,0.6964285714285714,0.4642857142857143,3.892857142857143,0.5357142857142857,56,0.5357142857142857,37.9,java,0.0
+5.011095568495564,0.5454545454545454,0.32727272727272727,5.490909090909091,0.6727272727272727,55,0.6727272727272727,33.93333333333333,java,1.0
+4.07744157582188,0.8085106382978723,0.5531914893617021,2.1702127659574466,0.44680851063829785,47,0.44680851063829785,31.142857142857142,python,0.0
+4.0614821867207755,0.3181818181818182,0.18181818181818177,3.8181818181818183,0.8181818181818182,22,0.8181818181818182,29.75,python,1.0
+4.490765292824625,0.7254901960784313,0.47058823529411764,3.019607843137255,0.5294117647058824,51,0.5294117647058824,22.727272727272727,python,1.0
+3.7695855752178566,0.9612403100775194,0.8372093023255813,2.550387596899225,0.16279069767441862,129,0.16279069767441862,26.833333333333332,java,0.0
+5.113304524601207,0.8288288288288288,0.5945945945945945,3.936936936936937,0.40540540540540543,111,0.40540540540540543,26.9375,java,0.0
+6.683599316096754,0.8333333333333334,0.6434599156118144,5.89662447257384,0.35654008438818563,474,0.35654008438818563,44.127906976744185,python,0.0
+5.742479746741097,0.8826291079812206,0.6572769953051643,7.413145539906103,0.3427230046948357,213,0.3427230046948357,34.11538461538461,java,1.0
+5.684220757851569,0.842809364548495,0.6889632107023411,4.11371237458194,0.3110367892976589,299,0.3110367892976589,42.95348837209303,python,0.0
+3.7248747399889837,0.8125,0.6041666666666667,3.6666666666666665,0.3958333333333333,48,0.3958333333333333,26.857142857142858,python,0.0
+4.438721875540867,0.7083333333333334,0.45833333333333337,3.5625,0.5416666666666666,48,0.5416666666666666,14.647058823529411,java,0.0
+5.052330376272108,0.7604166666666666,0.5416666666666667,5.239583333333333,0.4583333333333333,96,0.4583333333333333,37.76190476190476,java,0.0
+4.039211536948156,0.5161290322580645,0.3548387096774194,3.3870967741935485,0.6451612903225806,31,0.6451612903225806,12.0,java,0.0
+6.058508660286119,0.8492307692307692,0.6830769230769231,6.8246153846153845,0.3169230769230769,325,0.3169230769230769,40.19047619047619,java,1.0
+4.85654598061365,0.8559322033898306,0.6440677966101696,3.610169491525424,0.3559322033898305,118,0.3559322033898305,22.06896551724138,java,0.0
+4.384791749673746,0.8899082568807339,0.7155963302752293,2.6055045871559632,0.28440366972477066,109,0.28440366972477066,33.44444444444444,java,1.0
+4.2115044085684765,0.631578947368421,0.42105263157894735,4.473684210526316,0.5789473684210527,38,0.5789473684210527,31.77777777777778,python,1.0
+6.445861001406715,0.7741935483870968,0.5443548387096775,5.548387096774194,0.45564516129032256,248,0.45564516129032256,36.03508771929825,python,0.0
+4.060861683988659,0.8043478260869565,0.5652173913043479,3.0,0.43478260869565216,46,0.43478260869565216,28.77777777777778,python,1.0
+4.770281067729926,0.6226415094339622,0.39622641509433965,4.962264150943396,0.6037735849056604,53,0.6037735849056604,29.642857142857142,java,1.0
+4.310866879118227,0.8829787234042553,0.6914893617021276,4.340425531914893,0.30851063829787234,94,0.30851063829787234,30.791666666666668,python,0.0
+4.912733576964344,0.5636363636363636,0.36363636363636365,3.4545454545454546,0.6363636363636364,55,0.6363636363636364,24.705882352941178,python,1.0
+3.76062054430996,0.6666666666666666,0.48484848484848486,3.8484848484848486,0.5151515151515151,33,0.5151515151515151,27.11111111111111,python,1.0
+5.092277224236287,0.6621621621621622,0.43243243243243246,3.7837837837837838,0.5675675675675675,74,0.5675675675675675,29.5625,java,0.0
+3.5,0.5,0.25,3.125,0.75,16,0.75,19.75,python,1.0
+4.62765192370225,0.952,0.736,2.472,0.264,125,0.264,35.714285714285715,python,0.0
+4.516221717727311,0.8026315789473685,0.6052631578947368,5.197368421052632,0.39473684210526316,76,0.39473684210526316,27.72222222222222,python,0.0
+3.937751955408587,0.6451612903225806,0.4193548387096774,4.193548387096774,0.5806451612903226,31,0.5806451612903226,29.25,python,0.0
+5.000896232556761,0.8045977011494253,0.5402298850574713,3.6091954022988504,0.45977011494252873,87,0.45977011494252873,33.0,python,1.0
+6.038322688621861,0.827906976744186,0.6046511627906976,6.120930232558139,0.3953488372093023,215,0.3953488372093023,49.08108108108108,python,0.0
+4.216152190789069,0.8163265306122449,0.5510204081632653,3.5714285714285716,0.4489795918367347,49,0.4489795918367347,22.263157894736842,java,0.0
+4.667187378614269,0.8235294117647058,0.5882352941176471,4.0588235294117645,0.4117647058823529,85,0.4117647058823529,32.857142857142854,java,0.0
+5.070332970220562,0.8478260869565217,0.6594202898550725,4.297101449275362,0.34057971014492755,138,0.34057971014492755,30.34285714285714,java,0.0
+4.2366344310623845,0.9058823529411765,0.7058823529411764,3.8941176470588235,0.29411764705882354,85,0.29411764705882354,35.0,java,1.0
+4.062382847947341,0.5454545454545454,0.36363636363636365,2.393939393939394,0.6363636363636364,33,0.6363636363636364,24.333333333333332,python,1.0
+4.477226049206228,0.6851851851851852,0.4629629629629629,3.8703703703703702,0.5370370370370371,54,0.5370370370370371,37.0,java,1.0
+4.189125478116705,0.6521739130434783,0.4782608695652174,5.217391304347826,0.5217391304347826,46,0.5217391304347826,39.77777777777778,java,1.0
+3.6818808028034025,0.47368421052631576,0.26315789473684215,3.3157894736842106,0.7368421052631579,19,0.7368421052631579,22.75,python,1.0
+4.183768266288633,0.78125,0.609375,1.96875,0.390625,64,0.390625,23.25,python,1.0
+4.063202513215852,0.7368421052631579,0.4736842105263158,3.1052631578947367,0.5263157894736842,38,0.5263157894736842,23.77777777777778,python,1.0
+5.240360496391271,0.7901234567901234,0.4814814814814815,4.592592592592593,0.5185185185185185,81,0.5185185185185185,35.0,java,0.0
+2.9870900416203345,0.7619047619047619,0.5238095238095238,4.333333333333333,0.47619047619047616,21,0.47619047619047616,18.444444444444443,python,1.0
+4.796230935589923,0.917910447761194,0.7238805970149254,3.1194029850746268,0.27611940298507465,134,0.27611940298507465,33.964285714285715,java,0.0
+4.921133626426239,0.6779661016949152,0.4067796610169492,4.576271186440678,0.5932203389830508,59,0.5932203389830508,37.9,python,0.0
+5.223430137239087,0.9595375722543352,0.838150289017341,3.3757225433526012,0.16184971098265896,346,0.16184971098265896,26.730337078651687,java,0.0
+5.035549713367317,0.5454545454545454,0.32727272727272727,3.7636363636363637,0.6727272727272727,55,0.6727272727272727,25.272727272727273,python,0.0
+5.404449012034224,0.7906976744186046,0.5658914728682171,5.658914728682171,0.43410852713178294,129,0.43410852713178294,36.758620689655174,java,0.0
+4.629412138272088,0.4418604651162791,0.3023255813953488,3.8372093023255816,0.6976744186046512,43,0.6976744186046512,27.7,python,0.0
+4.816777664828826,0.9296875,0.71875,4.7109375,0.28125,128,0.28125,37.078947368421055,java,0.0
+5.5887549870381,0.8138297872340425,0.6276595744680851,4.388297872340425,0.3723404255319149,188,0.3723404255319149,50.24,python,0.0
+4.453150309616965,0.8518518518518519,0.7185185185185186,2.4074074074074074,0.2814814814814815,135,0.2814814814814815,14.0,java,0.0
+5.547057692620618,0.9096385542168675,0.7710843373493976,4.421686746987952,0.2289156626506024,332,0.2289156626506024,31.275862068965516,java,0.0
+4.627801780853476,0.9,0.6555555555555556,5.055555555555555,0.34444444444444444,90,0.34444444444444444,34.04761904761905,java,0.0
+4.747191429566854,0.5333333333333333,0.3111111111111111,5.466666666666667,0.6888888888888889,45,0.6888888888888889,36.44444444444444,python,0.0
+5.524469760991255,0.6788990825688074,0.45871559633027525,4.91743119266055,0.5412844036697247,109,0.5412844036697247,53.38461538461539,python,0.0
+5.347434988263512,0.4444444444444444,0.29166666666666663,4.0,0.7083333333333334,72,0.7083333333333334,45.625,python,0.0
+4.074705922236444,0.847457627118644,0.6271186440677966,3.6779661016949152,0.3728813559322034,59,0.3728813559322034,23.1,python,1.0
+4.41506101220307,0.43333333333333335,0.23333333333333328,4.833333333333333,0.7666666666666667,30,0.7666666666666667,17.333333333333332,python,0.0
+5.489282954487391,0.7518796992481203,0.5488721804511278,6.7443609022556394,0.45112781954887216,133,0.45112781954887216,41.4,java,0.0
+4.3199650731621695,0.8584070796460177,0.7256637168141593,3.3628318584070795,0.2743362831858407,113,0.2743362831858407,26.551724137931036,python,0.0
+5.948852612281259,0.8584905660377359,0.6886792452830188,6.908805031446541,0.3113207547169811,318,0.3113207547169811,39.423529411764704,java,1.0
+4.251365544851387,0.957983193277311,0.773109243697479,2.302521008403361,0.226890756302521,119,0.226890756302521,25.958333333333332,python,1.0
+4.966534113288597,0.6,0.41538461538461535,3.753846153846154,0.5846153846153846,65,0.5846153846153846,34.45454545454545,python,1.0
+4.216421766792934,0.6808510638297872,0.4893617021276596,3.4468085106382977,0.5106382978723404,47,0.5106382978723404,25.53846153846154,python,1.0
+5.854429565880467,0.47619047619047616,0.32380952380952377,5.79047619047619,0.6761904761904762,105,0.6761904761904762,32.892857142857146,python,1.0
+5.542084533777365,0.8870967741935484,0.7217741935483871,4.866935483870968,0.2782258064516129,248,0.2782258064516129,38.97727272727273,java,0.0
+3.8370959941879037,0.7560975609756098,0.5609756097560976,2.5853658536585367,0.43902439024390244,41,0.43902439024390244,23.8,python,1.0
+5.510000244558343,0.8188405797101449,0.572463768115942,5.884057971014493,0.427536231884058,138,0.427536231884058,31.333333333333332,java,0.0
+3.9171736218037974,0.9090909090909091,0.7121212121212122,2.878787878787879,0.2878787878787879,66,0.2878787878787879,30.307692307692307,java,1.0
+4.955593373471456,0.9126984126984127,0.6904761904761905,2.7222222222222223,0.30952380952380953,126,0.30952380952380953,31.208333333333332,python,1.0
+4.123966016311361,0.7407407407407407,0.537037037037037,3.074074074074074,0.46296296296296297,54,0.46296296296296297,29.9,python,0.0
+5.280191443467718,0.7971014492753623,0.6086956521739131,4.173913043478261,0.391304347826087,138,0.391304347826087,21.972972972972972,java,0.0
+4.616874605956221,0.47058823529411764,0.23529411764705888,5.147058823529412,0.7647058823529411,34,0.7647058823529411,33.142857142857146,python,1.0
+4.128326967635621,0.8043478260869565,0.5434782608695652,4.108695652173913,0.45652173913043476,46,0.45652173913043476,30.363636363636363,java,1.0
+4.589416279148799,0.8409090909090909,0.6363636363636364,2.715909090909091,0.36363636363636365,88,0.36363636363636365,17.785714285714285,java,0.0
+4.935672972133341,0.9946236559139785,0.8064516129032258,4.091397849462366,0.1935483870967742,186,0.1935483870967742,31.8,java,0.0
+4.76880107599223,0.8446601941747572,0.6310679611650485,3.8446601941747574,0.36893203883495146,103,0.36893203883495146,52.72727272727273,python,0.0
+4.1699823363507,0.8478260869565217,0.5434782608695652,5.456521739130435,0.45652173913043476,46,0.45652173913043476,33.45454545454545,java,1.0
+4.017917900762096,0.7058823529411765,0.4411764705882353,4.117647058823529,0.5588235294117647,34,0.5588235294117647,30.571428571428573,python,0.0
+4.6586089678214995,0.896,0.704,2.696,0.296,125,0.296,31.40740740740741,python,0.0
+4.643160624658742,0.7258064516129032,0.5,3.3870967741935485,0.5,62,0.5,45.30769230769231,java,0.0
+4.813012081410429,0.7857142857142857,0.5285714285714286,4.257142857142857,0.4714285714285714,70,0.4714285714285714,31.647058823529413,python,1.0
+4.60865917212318,0.9142857142857143,0.6952380952380952,2.6952380952380954,0.3047619047619048,105,0.3047619047619048,28.5,python,1.0
+4.713243393050852,0.9347826086956522,0.7536231884057971,2.152173913043478,0.2463768115942029,138,0.2463768115942029,33.76190476190476,python,0.0
+4.619054056131191,0.734375,0.515625,3.296875,0.484375,64,0.484375,28.066666666666666,java,0.0
+4.825245826488317,0.7654320987654321,0.5308641975308642,4.580246913580247,0.4691358024691358,81,0.4691358024691358,33.9,java,0.0
+6.527235468349858,0.8012232415902141,0.6085626911314985,5.792048929663609,0.39143730886850153,327,0.39143730886850153,42.516129032258064,python,0.0
+5.59181229230442,0.7133333333333334,0.5333333333333333,4.233333333333333,0.4666666666666667,150,0.4666666666666667,33.55172413793103,python,0.0
+4.0006374384002905,0.8421052631578947,0.631578947368421,2.491228070175439,0.3684210526315789,57,0.3684210526315789,33.5,python,1.0
+3.7461571189724303,0.9545454545454546,0.6590909090909092,4.25,0.3409090909090909,44,0.3409090909090909,28.833333333333332,python,1.0
+5.745512973329542,0.8275862068965517,0.5517241379310345,6.027586206896552,0.4482758620689655,145,0.4482758620689655,34.648648648648646,java,0.0
+5.490266518243658,0.8412698412698413,0.6243386243386244,5.121693121693122,0.37566137566137564,189,0.37566137566137564,40.166666666666664,java,0.0
+4.496115365169272,0.6875,0.4375,4.729166666666667,0.5625,48,0.5625,43.285714285714285,python,1.0
+5.505436888624101,0.945273631840796,0.7164179104477613,3.855721393034826,0.2835820895522388,201,0.2835820895522388,35.58139534883721,java,0.0
+4.795079572399797,0.8155339805825242,0.6116504854368932,4.621359223300971,0.3883495145631068,103,0.3883495145631068,37.54545454545455,java,0.0
+5.595002345872928,0.7768595041322314,0.5206611570247934,6.099173553719008,0.4793388429752066,121,0.4793388429752066,32.029411764705884,java,0.0
+5.887085632611005,0.93359375,0.68359375,6.109375,0.31640625,256,0.31640625,31.675675675675677,java,0.0
+4.809267348202161,0.7971014492753623,0.5217391304347826,3.8840579710144927,0.4782608695652174,69,0.4782608695652174,38.15384615384615,python,0.0
+4.893291230757242,0.5737704918032787,0.39344262295081966,4.311475409836065,0.6065573770491803,61,0.6065573770491803,33.7,python,0.0
+4.346926022805004,0.8,0.5454545454545454,4.072727272727272,0.45454545454545453,55,0.45454545454545453,29.866666666666667,java,0.0
+4.476474118254376,0.7857142857142857,0.5178571428571428,6.660714285714286,0.48214285714285715,56,0.48214285714285715,39.92307692307692,java,1.0
+3.9809099797950775,0.9154929577464789,0.704225352112676,4.422535211267606,0.29577464788732394,71,0.29577464788732394,30.0,java,1.0
+5.272508205443481,0.5967741935483871,0.32258064516129037,6.193548387096774,0.6774193548387096,62,0.6774193548387096,44.166666666666664,java,1.0
+5.1902361166256314,0.9676113360323887,0.7854251012145749,4.611336032388664,0.2145748987854251,247,0.2145748987854251,28.724137931034484,java,1.0
+4.350489295854283,0.7333333333333333,0.55,3.183333333333333,0.45,60,0.45,33.8,java,1.0
+5.0009504388098796,0.7927927927927928,0.5765765765765766,4.5225225225225225,0.42342342342342343,111,0.42342342342342343,35.142857142857146,java,0.0
+5.7769401731183265,0.9180327868852459,0.6284153005464481,5.573770491803279,0.37158469945355194,183,0.37158469945355194,28.34375,java,0.0
+3.876189125313799,0.8970588235294118,0.6911764705882353,2.073529411764706,0.3088235294117647,68,0.3088235294117647,27.9,java,1.0
+4.032303242743952,0.4444444444444444,0.2962962962962963,3.888888888888889,0.7037037037037037,27,0.7037037037037037,26.5,python,0.0
+3.625,0.375,0.1875,6.625,0.8125,16,0.8125,27.8,python,1.0
+4.4210998675908275,0.9540816326530612,0.826530612244898,2.4846938775510203,0.17346938775510204,196,0.17346938775510204,35.06896551724138,python,0.0
+4.839775539645508,0.3157894736842105,0.1842105263157895,5.157894736842105,0.8157894736842105,38,0.8157894736842105,37.0,python,1.0
+5.842252978929465,0.8711484593837535,0.7254901960784313,6.647058823529412,0.27450980392156865,357,0.27450980392156865,42.16470588235294,java,1.0
+4.166886761135438,0.78,0.56,3.2,0.44,50,0.44,29.916666666666668,python,0.0
+5.395683189409616,0.8947368421052632,0.6947368421052631,5.515789473684211,0.30526315789473685,190,0.30526315789473685,29.24590163934426,java,1.0

featureextraction/step1_statistical_extraction/val_features.csv

@@ -0,0 +1,75 @@
+entropy,burstiness,repetition_ratio,avg_token_length,vocab_richness,num_tokens,unique_token_ratio,avg_line_length,language,Label
+3.616348566075164,0.42105263157894735,0.26315789473684215,3.4210526315789473,0.7368421052631579,19,0.7368421052631579,25.25,python,1.0
+5.225116096801971,0.9219512195121952,0.7365853658536585,4.84390243902439,0.2634146341463415,205,0.2634146341463415,31.340425531914892,java,0.0
+5.858683407719383,0.5053763440860215,0.30107526881720426,6.021505376344086,0.6989247311827957,93,0.6989247311827957,30.59259259259259,python,1.0
+3.4687406100460665,0.918918918918919,0.6486486486486487,2.4324324324324325,0.35135135135135137,37,0.35135135135135137,40.333333333333336,python,1.0
+4.726474118254375,0.75,0.4642857142857143,6.339285714285714,0.5357142857142857,56,0.5357142857142857,33.25,java,1.0
+5.127518057233353,0.9186602870813397,0.7416267942583732,4.779904306220096,0.2583732057416268,209,0.2583732057416268,42.84848484848485,java,0.0
+3.385592490318422,0.8181818181818182,0.6909090909090909,2.6,0.3090909090909091,55,0.3090909090909091,23.75,python,1.0
+5.303371529748239,0.8614457831325302,0.6566265060240963,4.445783132530121,0.3433734939759036,166,0.3433734939759036,19.36,java,0.0
+4.32307418942857,0.72,0.5,4.04,0.5,50,0.5,23.647058823529413,python,1.0
+4.646104113627779,0.6666666666666666,0.4444444444444444,3.925925925925926,0.5555555555555556,54,0.5555555555555556,17.8125,java,0.0
+4.824043435284101,0.7971014492753623,0.5217391304347826,3.9130434782608696,0.4782608695652174,69,0.4782608695652174,26.157894736842106,python,0.0
+4.133660689688186,0.44,0.24,6.0,0.76,25,0.76,28.625,java,1.0
+3.977232016069148,0.8854166666666666,0.7395833333333333,2.7395833333333335,0.2604166666666667,96,0.2604166666666667,28.4375,python,1.0
+4.418945246473782,0.8461538461538461,0.641025641025641,3.1153846153846154,0.358974358974359,78,0.358974358974359,33.78947368421053,java,0.0
+3.795088586397732,0.4782608695652174,0.30434782608695654,3.5652173913043477,0.6956521739130435,23,0.6956521739130435,21.166666666666668,python,1.0
+4.665893393157604,0.8735632183908046,0.632183908045977,2.4942528735632186,0.367816091954023,87,0.367816091954023,30.833333333333332,java,1.0
+4.888279432092592,0.5636363636363636,0.36363636363636365,4.5636363636363635,0.6363636363636364,55,0.6363636363636364,31.7,python,0.0
+4.43887358554625,0.7111111111111111,0.4444444444444444,4.488888888888889,0.5555555555555556,45,0.5555555555555556,27.533333333333335,java,0.0
+4.956005194695069,0.7636363636363637,0.5909090909090908,3.963636363636364,0.4090909090909091,110,0.4090909090909091,22.925925925925927,java,0.0
+4.364611584126849,0.8990825688073395,0.7201834862385321,3.86697247706422,0.2798165137614679,218,0.2798165137614679,36.03125,python,0.0
+5.1310450521201885,0.6923076923076923,0.4945054945054945,3.8241758241758244,0.5054945054945055,91,0.5054945054945055,43.083333333333336,python,0.0
+4.157269582127642,0.9487179487179487,0.7863247863247863,1.735042735042735,0.21367521367521367,117,0.21367521367521367,36.583333333333336,python,1.0
+6.1032071909268755,0.7524271844660194,0.558252427184466,4.344660194174757,0.441747572815534,206,0.441747572815534,42.59375,python,0.0
+5.489504669651612,0.9067796610169492,0.7203389830508475,3.983050847457627,0.2796610169491525,236,0.2796610169491525,46.03225806451613,java,0.0
+5.5429756373358305,0.6216216216216216,0.4504504504504504,3.810810810810811,0.5495495495495496,111,0.5495495495495496,46.0,python,0.0
+5.597773058187228,0.7096774193548387,0.4193548387096774,4.548387096774194,0.5806451612903226,93,0.5806451612903226,32.27777777777778,python,1.0
+4.288320189106,0.7916666666666666,0.5208333333333333,4.229166666666667,0.4791666666666667,48,0.4791666666666667,34.0,python,0.0
+5.117143809076922,0.8490566037735849,0.5849056603773585,3.943396226415094,0.41509433962264153,106,0.41509433962264153,39.4,python,0.0
+4.3341837197791895,0.8,0.44999999999999996,7.05,0.55,40,0.55,41.666666666666664,java,1.0
+4.7319988769018755,0.8416666666666667,0.675,3.091666666666667,0.325,120,0.325,15.944444444444445,java,0.0
+4.2628148954723555,0.55,0.4,2.825,0.6,40,0.6,31.285714285714285,python,0.0
+3.25,0.875,0.625,3.6875,0.375,32,0.375,41.4,java,1.0
+5.086021506160257,0.7008547008547008,0.5470085470085471,4.239316239316239,0.452991452991453,117,0.452991452991453,44.125,python,0.0
+4.488525294897748,0.6923076923076923,0.46153846153846156,3.9038461538461537,0.5384615384615384,52,0.5384615384615384,30.181818181818183,java,1.0
+5.273684376262023,0.6111111111111112,0.375,4.819444444444445,0.625,72,0.625,42.72727272727273,python,0.0
+4.047055675509121,0.8627450980392157,0.607843137254902,4.372549019607843,0.39215686274509803,51,0.39215686274509803,28.428571428571427,python,1.0
+4.6442547530701885,0.5370370370370371,0.38888888888888884,5.12962962962963,0.6111111111111112,54,0.6111111111111112,34.7,java,0.0
+4.423251796980337,0.32142857142857145,0.1785714285714286,2.642857142857143,0.8214285714285714,28,0.8214285714285714,30.0,python,1.0
+5.498927847038898,0.5842696629213483,0.3820224719101124,6.617977528089888,0.6179775280898876,89,0.6179775280898876,41.421052631578945,python,1.0
+4.376857192109841,0.8625,0.65,3.4125,0.35,80,0.35,29.764705882352942,java,0.0
+4.600176636857892,0.7894736842105263,0.5087719298245614,3.6842105263157894,0.49122807017543857,57,0.49122807017543857,30.5,python,1.0
+6.213552119915352,0.8066037735849056,0.5613207547169812,5.485849056603773,0.4386792452830189,212,0.4386792452830189,39.6,python,0.0
+4.024059741690056,0.7727272727272727,0.6060606060606061,3.5454545454545454,0.3939393939393939,66,0.3939393939393939,30.46153846153846,python,1.0
+4.14853573331382,0.7619047619047619,0.5,7.023809523809524,0.5,42,0.5,26.266666666666666,java,0.0
+4.2057539822273915,0.8507462686567164,0.6417910447761195,3.5522388059701493,0.3582089552238806,67,0.3582089552238806,30.352941176470587,python,0.0
+4.860498585099573,0.8737864077669902,0.6504854368932038,4.097087378640777,0.34951456310679613,103,0.34951456310679613,32.61538461538461,java,0.0
+4.661648023061636,0.7631578947368421,0.5526315789473684,3.776315789473684,0.4473684210526316,76,0.4473684210526316,32.05555555555556,java,0.0
+5.405165733375986,0.9311926605504587,0.7339449541284404,5.426605504587156,0.26605504587155965,218,0.26605504587155965,33.0,java,0.0
+3.238901256602631,0.46153846153846156,0.23076923076923073,4.538461538461538,0.7692307692307693,13,0.7692307692307693,26.0,python,1.0
+3.964395529526285,0.9512195121951219,0.5853658536585367,4.170731707317073,0.4146341463414634,41,0.4146341463414634,27.1,java,1.0
+4.091031702849688,0.8170731707317073,0.6707317073170731,3.8780487804878048,0.32926829268292684,82,0.32926829268292684,31.833333333333332,python,0.0
+3.6596769481795466,0.8571428571428571,0.6190476190476191,4.976190476190476,0.38095238095238093,42,0.38095238095238093,24.533333333333335,java,1.0
+5.249145047994913,0.6875,0.5178571428571428,3.8125,0.48214285714285715,112,0.48214285714285715,46.5,python,0.0
+4.456796057875661,0.7761194029850746,0.582089552238806,2.283582089552239,0.417910447761194,67,0.417910447761194,25.545454545454547,java,1.0
+5.669233537529181,0.8121827411167513,0.6142131979695431,4.3908629441624365,0.38578680203045684,197,0.38578680203045684,49.56,python,0.0
+4.923879361943277,0.9572649572649573,0.6923076923076923,4.205128205128205,0.3076923076923077,117,0.3076923076923077,31.321428571428573,python,0.0
+4.028579035079149,0.9841269841269841,0.7142857142857143,3.8253968253968256,0.2857142857142857,63,0.2857142857142857,32.666666666666664,python,1.0
+4.931270849848961,0.6,0.36363636363636365,5.763636363636364,0.6363636363636364,55,0.6363636363636364,29.8125,java,1.0
+4.987993396430413,0.8682170542635659,0.6744186046511628,3.317829457364341,0.32558139534883723,129,0.32558139534883723,33.02777777777778,java,0.0
+4.895499903544709,0.9303482587064676,0.7661691542288558,3.4776119402985075,0.23383084577114427,201,0.23383084577114427,23.032258064516128,java,0.0
+4.297689964750703,0.9158878504672897,0.7476635514018692,3.532710280373832,0.2523364485981308,107,0.2523364485981308,31.107142857142858,java,0.0
+4.531740514375559,0.8,0.509090909090909,5.9818181818181815,0.4909090909090909,55,0.4909090909090909,30.25,java,1.0
+4.894400915540714,0.7763157894736842,0.5263157894736843,3.3157894736842106,0.47368421052631576,76,0.47368421052631576,29.952380952380953,python,0.0
+5.754451229096238,0.7107438016528925,0.4628099173553719,5.603305785123967,0.5371900826446281,121,0.5371900826446281,34.03703703703704,java,0.0
+3.5141016549301787,0.6944444444444444,0.5555555555555556,3.0277777777777777,0.4444444444444444,36,0.4444444444444444,33.6,python,1.0
+5.4389854001411315,0.7350427350427351,0.5213675213675213,5.170940170940171,0.47863247863247865,117,0.47863247863247865,43.72727272727273,java,0.0
+6.500985818092827,0.8327974276527331,0.6109324758842444,6.864951768488746,0.3890675241157556,311,0.3890675241157556,38.79775280898876,java,1.0
+3.9219280948873627,0.4,0.19999999999999996,3.85,0.8,20,0.8,20.5,python,1.0
+4.594928174344042,0.9072164948453608,0.6804123711340206,2.6804123711340204,0.31958762886597936,97,0.31958762886597936,35.5,java,1.0
+6.3478254862783485,0.7662337662337663,0.5454545454545454,6.783549783549783,0.45454545454545453,231,0.45454545454545453,33.81818181818182,java,1.0
+4.777798184597433,0.8349514563106796,0.6310679611650485,4.475728155339806,0.36893203883495146,103,0.36893203883495146,30.413793103448278,python,0.0
+4.955482172039427,0.7422680412371134,0.5257731958762887,5.556701030927835,0.4742268041237113,97,0.4742268041237113,29.88,python,0.0
+3.963119345888566,0.7291666666666666,0.5416666666666667,3.8333333333333335,0.4583333333333333,48,0.4583333333333333,29.2,python,1.0
+4.981129381113054,0.7857142857142857,0.5892857142857143,3.8214285714285716,0.4107142857142857,112,0.4107142857142857,47.69230769230769,python,0.0