Dataset Viewer
Auto-converted to Parquet
Date
string
india_vix
float64
rsi_14
float64
ma_50
float64
ma_200
float64
regime
int64
2015-10-23
8,295.450195
34.61122
20.6624
17.89865
1
2015-10-26
8,260.549805
41.437422
20.6892
17.9158
0
2015-10-27
8,232.900391
38.561175
20.671
17.92775
1
2015-10-28
8,171.200195
41.718164
20.6714
17.92585
1
2015-10-29
8,111.75
45.004878
20.7026
17.92305
0
2015-10-30
8,065.799805
46.799834
20.731
17.9301
0
2015-11-02
8,050.799805
55.19844
20.7934
17.9477
0
2015-11-03
8,060.700195
55.576069
20.851
17.96495
0
2015-11-04
8,040.200195
54.342858
20.8986
17.97935
0
2015-11-05
7,955.450195
53.892116
20.9422
17.98955
0
2015-11-06
7,954.299805
54.92739
20.769
18.0061
0
2015-11-09
7,915.200195
41.409655
20.5728
18.00495
0
2015-11-10
7,783.350098
40.482438
20.399
18.00085
0
2015-11-13
7,762.25
45.604233
20.3112
18.00275
0
2015-11-16
7,806.600098
46.982379
20.2022
18.0036
0
2015-11-17
7,837.549805
42.40472
20.0496
17.9956
0
2015-11-18
7,731.799805
45.245794
19.8236
17.99325
0
2015-11-19
7,842.75
38.133209
19.6074
17.982
1
2015-11-20
7,856.549805
38.528393
19.443
17.96265
1
2015-11-23
7,849.25
45.853753
19.2562
17.9509
0
2015-11-24
7,831.600098
42.057313
19.0542
17.9313
1
2015-11-26
7,883.799805
44.672499
18.8966
17.9127
1
2015-11-27
7,942.700195
44.915893
18.7452
17.8973
1
2015-11-30
7,935.25
43.374683
18.5788
17.8792
1
2015-12-01
7,954.899902
39.528228
18.392
17.8543
1
2015-12-02
7,931.350098
38.82422
18.2154
17.8283
1
2015-12-03
7,864.149902
42.980101
18.0618
17.7985
1
2015-12-04
7,781.899902
43.198818
17.9436
17.773
1
2015-12-07
7,765.399902
41.928701
17.8952
17.74985
1
2015-12-08
7,701.700195
42.752631
17.851
17.7289
1
2015-12-09
7,612.5
48.708815
17.7786
17.71205
1
2015-12-10
7,683.299805
45.741615
17.6906
17.68935
1
2015-12-11
7,610.450195
52.098075
17.6166
17.6707
1
2015-12-14
7,650.049805
55.968003
17.5392
17.6577
1
2015-12-15
7,700.899902
52.574298
17.4546
17.63805
1
2015-12-16
7,750.899902
48.384452
17.3966
17.61355
1
2015-12-17
7,844.350098
35.757045
17.2962
17.5769
1
2015-12-18
7,761.950195
37.64231
17.2034
17.54545
1
2015-12-21
7,834.450195
36.530581
17.1018
17.5137
1
2015-12-22
7,786.100098
38.414037
17.0046
17.48825
1
2015-12-23
7,865.950195
34.729276
16.8864
17.47725
1
2015-12-24
7,861.049805
35.531645
16.7828
17.46855
1
2015-12-28
7,925.149902
40.639618
16.6884
17.46455
1
2015-12-29
7,928.950195
39.803601
16.6128
17.46355
1
2015-12-30
7,896.25
40.706692
16.5454
17.45625
1
2015-12-31
7,946.350098
38.025849
16.4794
17.44795
1
2016-01-01
7,946.350098
41.402548
16.4392
17.4436
1
2016-01-04
7,791.299805
57.787964
16.4384
17.45405
1
2016-01-05
7,784.649902
56.858853
16.4324
17.4627
1
2016-01-06
7,741
55.823234
16.431
17.46955
1
2016-01-07
7,568.299805
66.409155
16.4864
17.4887
1
2016-01-08
7,601.350098
59.071785
16.4978
17.50015
1
2016-01-11
7,563.850098
62.495613
16.5408
17.5176
1
2016-01-12
7,510.299805
62.571381
16.5742
17.53985
1
2016-01-13
7,562.399902
60.213173
16.5896
17.56025
1
2016-01-14
7,536.799805
61.34624
16.6042
17.58495
1
2016-01-15
7,437.799805
64.717351
16.603
17.61635
1
2016-01-18
7,351
67.746642
16.6164
17.6417
0
2016-01-19
7,435.100098
56.19928
16.5978
17.66135
1
2016-01-20
7,309.299805
65.210591
16.6314
17.6949
0
2016-01-21
7,276.799805
64.313004
16.6582
17.7265
0
2016-01-22
7,422.450195
54.301719
16.6938
17.75225
0
2016-01-25
7,436.149902
51.897198
16.7226
17.77075
0
2016-01-27
7,437.75
54.980824
16.7508
17.7948
0
2016-01-28
7,424.649902
49.920415
16.7512
17.81255
0
2016-01-29
7,563.549805
47.255419
16.7566
17.82665
0
2016-02-01
7,555.950195
50.081803
16.766
17.8448
0
2016-02-02
7,455.549805
50.737849
16.81
17.86255
0
2016-02-03
7,361.799805
53.16819
16.8638
17.8824
0
2016-02-04
7,404
51.285596
16.8864
17.8987
0
2016-02-05
7,489.100098
50.776081
16.9236
17.91405
0
2016-02-08
7,387.25
59.98165
16.995
17.93385
0
2016-02-09
7,298.200195
64.825256
17.095
17.95735
0
2016-02-10
7,215.700195
66.99116
17.2158
17.98365
0
2016-02-11
6,976.350098
74.872202
17.4224
18.021
0
2016-02-12
6,980.950195
68.345098
17.606
18.0488
0
2016-02-15
7,162.950195
58.646303
17.7324
18.067
0
2016-02-16
7,048.25
60.390917
17.8714
18.09515
0
2016-02-17
7,108.450195
55.747519
17.9906
18.1181
0
2016-02-18
7,191.75
54.479929
18.101
18.1394
0
2016-02-19
7,210.75
52.331616
18.1848
18.15745
0
2016-02-22
7,234.549805
52.684133
18.2794
18.17575
0
2016-02-23
7,109.549805
61.361721
18.4126
18.1966
0
2016-02-24
7,018.700195
57.79939
18.5158
18.21315
0
2016-02-25
6,970.600098
56.949725
18.624
18.23135
0
2016-02-26
7,029.75
53.406738
18.7268
18.24755
0
2016-02-29
6,987.049805
47.254665
18.8456
18.2456
0
2016-03-01
7,222.299805
42.645012
18.9292
18.2355
0
2016-03-02
7,368.850098
43.910486
19.0244
18.22695
0
2016-03-03
7,475.600098
41.017107
19.0956
18.21805
0
2016-03-04
7,485.350098
40.53268
19.1802
18.21795
0
2016-03-08
7,485.299805
43.196769
19.2746
18.21895
0
2016-03-09
7,531.799805
40.05261
19.3368
18.2179
0
2016-03-10
7,486.149902
41.902676
19.41
18.22035
0
2016-03-11
7,510.200195
39.228275
19.465
18.2211
0
2016-03-14
7,538.75
39.228275
19.5296
18.22195
0
2016-03-15
7,460.600098
41.915445
19.5964
18.2246
0
2016-03-16
7,498.75
40.046614
19.6018
18.22375
1
2016-03-17
7,512.549805
39.01439
19.6046
18.2223
1
2016-03-18
7,604.350098
37.25264
19.6012
18.2209
1
End of preview. Expand in Data Studio

πŸ“Š NIFTY 50 Market Regime Dataset

Dataset for training machine learning models to predict market regimes (RISK_ON/RISK_OFF) in Indian financial markets.

πŸ“‹ Dataset Description

This dataset contains technical indicators and corresponding market regime labels for NIFTY 50, used to train binary classification models for regime prediction.

Market Regimes

  • RISK_ON (1): Favorable market conditions - lower volatility, bullish momentum
  • RISK_OFF (0): Cautious conditions - higher volatility, defensive positioning

πŸ“Š Features

Feature Type Description Range
india_vix float India VIX volatility index 0-100+
rsi_14 float 14-day Relative Strength Index 0-100
ma_50 float 50-day moving average > 0
ma_200 float 200-day moving average > 0
regime int Target label (0=RISK_OFF, 1=RISK_ON) 0 or 1

Feature Descriptions

  1. India VIX: Measures expected volatility in NIFTY 50

    • Low values (< 15): Low fear, potentially RISK_ON
    • High values (> 25): High fear, potentially RISK_OFF
  2. RSI-14: Momentum indicator

    • < 30: Oversold (potentially bullish)
    • 70: Overbought (potentially bearish)

    • 40-60: Neutral
  3. MA-50: Short-term trend indicator

  4. MA-200: Long-term trend indicator

πŸ“ˆ Statistics

Load the dataset to see:

  • Total samples
  • Class distribution (RISK_ON vs RISK_OFF)
  • Feature correlations
  • Summary statistics

πŸ’» Usage

Load with Pandas

import pandas as pd

# Load from local file
df = pd.read_csv("market_regime_data_nifty50.csv")

# View first few rows
print(df.head())

# Check class distribution
print(df['regime'].value_counts())

Load from Hugging Face

from datasets import load_dataset

# Load dataset
dataset = load_dataset("AAdevloper/nifty50-market-regime")

# Convert to pandas
df = dataset['train'].to_pandas()

Example Training Code

from sklearn.model_selection import train_test_split
from xgboost import XGBClassifier

# Prepare data
X = df[['india_vix', 'rsi_14', 'ma_50', 'ma_200']]
y = df['regime']

# Split
X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=42, stratify=y
)

# Train
model = XGBClassifier(max_depth=6, learning_rate=0.1, n_estimators=100)
model.fit(X_train, y_train)

# Evaluate
score = model.score(X_test, y_test)
print(f"Accuracy: {score:.2%}")

🎯 Use Cases

  1. Market Regime Classification: Train models to predict current regime
  2. Trading Strategy Development: Use regime predictions for position sizing
  3. Risk Management: Adjust portfolio based on predicted regime
  4. Feature Engineering Research: Experiment with technical indicators
  5. MLOps Pipeline Development: Practice model deployment and monitoring

πŸ”§ Data Collection

The dataset includes:

  • Historical technical indicators for NIFTY 50
  • Calculated from OHLCV (Open, High, Low, Close, Volume) data
  • Regime labels based on market behavior patterns

πŸ“Š Data Quality

  • βœ… No missing values
  • βœ… Features normalized/scaled appropriately
  • βœ… Balanced class distribution (or document imbalance)
  • βœ… No data leakage in feature engineering
  • βœ… Temporally consistent

πŸ—οΈ Related Projects

This dataset is part of a complete MLOps pipeline:

πŸ“„ Citation

If you use this dataset, please cite:

@misc{nifty50_market_regime,
  author = {AAdevloper},
  title = {NIFTY 50 Market Regime Dataset},
  year = {2025},
  publisher = {Hugging Face},
  howpublished = {\url{https://huggingface.co/datasets/AAdevloper/nifty50-market-regime}}
}

πŸ“„ License

MIT License - Free to use for research and commercial applications


Part of the MLOps Finance Pipeline project πŸš€

For questions or improvements, visit the GitHub repository

Downloads last month
39