Spaces:
Running
Running
File size: 4,683 Bytes
9eecab5 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 | import json
from pathlib import Path
import pandas as pd
from utils.logger import logger
import os
DATASET_DIR = Path("data/datasets")
METADATA_DIR = Path("data/metadata")
DATASET_DIR.mkdir(parents=True, exist_ok=True)
METADATA_DIR.mkdir(parents=True, exist_ok=True)
class DatasetRegistry:
def __init__(self):
self.datasets = {}
self._load_existing()
def _load_existing(self):
try:
for meta_file in METADATA_DIR.glob("*.json"):
name = meta_file.stem
with open(meta_file, "r") as f:
metadata = json.load(f)
self.datasets[name] = metadata
logger.info(f"Loaded {len(self.datasets)} datasets into registry")
except Exception as e:
logger.error(f"Registry loading failed | {e}")
def delete_dataset(self, name):
try:
deleted_files = []
dataset_dir = "data/datasets"
metadata_dir = "data/metadata"
# Possible variations
variants = [name, f"{name}_clean"]
for variant in variants:
parquet_path = os.path.join(dataset_dir, f"{variant}.parquet")
metadata_path = os.path.join(metadata_dir, f"{variant}.json")
if os.path.exists(parquet_path):
os.remove(parquet_path)
deleted_files.append(parquet_path)
if os.path.exists(metadata_path):
os.remove(metadata_path)
deleted_files.append(metadata_path)
if not deleted_files:
return f"No dataset found for '{name}'"
logger.info(f"Deleted dataset {name} | Files: {deleted_files}")
return f"Deleted dataset '{name}' successfully."
except Exception as e:
logger.error(f"Delete failed | {e}")
return f"Failed to delete dataset '{name}'"
def register_dataset(self, name, df, schema):
try:
if name in self.datasets:
raise ValueError(f"Dataset '{name}' already exists")
parquet_path = DATASET_DIR / f"{name}.parquet"
meta_path = METADATA_DIR / f"{name}.json"
df.to_parquet(parquet_path)
with open(meta_path, "w") as f:
json.dump(schema, f, indent=2)
self.datasets[name] = schema
logger.info(f"Dataset registered | {name}")
except Exception as e:
logger.error(f"Dataset registration failed | {e}")
raise
def dataset_exists(self, name):
return name in self.datasets
def list_datasets(self):
return list(self.datasets.keys())
def get_info(self, name):
if name not in self.datasets:
raise ValueError("Dataset not found")
return self.datasets[name]
def update_dataset(self, name, df, schema):
try:
parquet_path = DATASET_DIR / f"{name}.parquet"
meta_path = METADATA_DIR / f"{name}.json"
df.to_parquet(parquet_path)
with open(meta_path, "w") as f:
json.dump(schema, f, indent=2)
self.datasets[name] = schema
logger.info(f"Dataset updated | {name}")
except Exception as e:
logger.error(f"Dataset update failed | {e}")
raise
def load_dataframe(self, name, sample=True, sample_size=50000):
try:
# ---------- VALIDATION ----------
if name not in self.datasets:
logger.error(f"Dataset '{name}' not found in registry")
raise ValueError(f"Dataset '{name}' not found")
path = DATASET_DIR / f"{name}.parquet"
if not path.exists():
logger.error(f"Parquet file missing: {path}")
raise FileNotFoundError(f"{path} not found")
logger.info(f"Loading dataset: {name}")
# ---------- LOAD ----------
df = pd.read_parquet(path)
logger.info(f"Loaded dataset '{name}' | shape={df.shape}")
# ---------- SMART SAMPLING ----------
if sample and len(df) > sample_size:
logger.info(
f"Dataset '{name}' is large ({len(df)} rows). "
f"Sampling {sample_size} rows for analysis."
)
df = df.sample(sample_size, random_state=42)
logger.info(f"Sampled dataset '{name}' | new_shape={df.shape}")
return df
except Exception as e:
logger.error(f"Failed to load dataset '{name}' | {e}")
raise |