File size: 5,899 Bytes
a74054f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
"""

Data cleaning utilities.

Single Responsibility: Clean and preprocess hydrological/meteorological data.

"""
import pandas as pd
import numpy as np
from typing import Optional, List


class DataCleaner:
    """

    Cleans hydrological and meteorological datasets.

    Handles missing values, outliers, and data quality issues.

    """

    def __init__(

        self,

        missing_threshold: float = 0.3,

        outlier_method: str = "iqr"

    ):
        """

        Initialize data cleaner.



        Args:

            missing_threshold: Max fraction of missing values allowed per column

            outlier_method: Method for outlier detection ("iqr", "zscore", "none")

        """
        self.missing_threshold = missing_threshold
        self.outlier_method = outlier_method

    def clean(

        self,

        df: pd.DataFrame,

        date_col: str = "date",

        station_col: str = "station_id"

    ) -> pd.DataFrame:
        """

        Clean dataframe with multiple steps.



        Args:

            df: Input dataframe

            date_col: Name of date column

            station_col: Name of station ID column



        Returns:

            Cleaned dataframe

        """
        df = df.copy()

        # Remove duplicates
        df = self._remove_duplicates(df, date_col, station_col)

        # Handle missing values
        df = self._handle_missing(df, date_col, station_col)

        # Remove outliers if requested
        if self.outlier_method != "none":
            df = self._remove_outliers(df, date_col, station_col)

        # Sort by date and station
        if date_col in df.columns and station_col in df.columns:
            df = df.sort_values([station_col, date_col]).reset_index(drop=True)

        return df

    def _remove_duplicates(

        self,

        df: pd.DataFrame,

        date_col: str,

        station_col: str

    ) -> pd.DataFrame:
        """Remove duplicate rows based on date and station."""
        if date_col in df.columns and station_col in df.columns:
            df = df.drop_duplicates(subset=[date_col, station_col], keep="first")
        return df

    def _handle_missing(

        self,

        df: pd.DataFrame,

        date_col: str,

        station_col: str

    ) -> pd.DataFrame:
        """

        Handle missing values.

        - Drop columns with >threshold missing

        - Forward fill small gaps in time series

        """
        # Drop columns with too many missing values
        missing_frac = df.isnull().mean()
        cols_to_keep = missing_frac[missing_frac <= self.missing_threshold].index
        df = df[cols_to_keep]

        # Forward fill small gaps (max 3 days) within each station
        if station_col in df.columns:
            numeric_cols = df.select_dtypes(include=[np.number]).columns
            df[numeric_cols] = df.groupby(station_col)[numeric_cols].transform(
                lambda x: x.fillna(method='ffill', limit=3)
            )

        return df

    def _remove_outliers(

        self,

        df: pd.DataFrame,

        date_col: str,

        station_col: str

    ) -> pd.DataFrame:
        """

        Remove outliers using IQR or Z-score method.

        Only applies to numeric columns.

        """
        numeric_cols = df.select_dtypes(include=[np.number]).columns
        numeric_cols = [c for c in numeric_cols if c != station_col]

        if self.outlier_method == "iqr":
            df = self._remove_outliers_iqr(df, numeric_cols, station_col)
        elif self.outlier_method == "zscore":
            df = self._remove_outliers_zscore(df, numeric_cols, station_col)

        return df

    def _remove_outliers_iqr(

        self,

        df: pd.DataFrame,

        cols: List[str],

        station_col: str

    ) -> pd.DataFrame:
        """Remove outliers using IQR method (per station)."""
        for col in cols:
            if col in df.columns:
                # Calculate IQR per station
                Q1 = df.groupby(station_col)[col].transform(lambda x: x.quantile(0.25))
                Q3 = df.groupby(station_col)[col].transform(lambda x: x.quantile(0.75))
                IQR = Q3 - Q1
                lower_bound = Q1 - 3 * IQR
                upper_bound = Q3 + 3 * IQR

                # Set outliers to NaN
                df.loc[(df[col] < lower_bound) | (df[col] > upper_bound), col] = np.nan

        return df

    def _remove_outliers_zscore(

        self,

        df: pd.DataFrame,

        cols: List[str],

        station_col: str,

        threshold: float = 4.0

    ) -> pd.DataFrame:
        """Remove outliers using Z-score method (per station)."""
        for col in cols:
            if col in df.columns:
                # Calculate Z-scores per station
                mean = df.groupby(station_col)[col].transform('mean')
                std = df.groupby(station_col)[col].transform('std')
                z_scores = np.abs((df[col] - mean) / std)

                # Set outliers to NaN
                df.loc[z_scores > threshold, col] = np.nan

        return df

    def get_cleaning_report(self, df_before: pd.DataFrame, df_after: pd.DataFrame) -> dict:
        """

        Generate report of cleaning operations.



        Args:

            df_before: DataFrame before cleaning

            df_after: DataFrame after cleaning



        Returns:

            Dictionary with cleaning statistics

        """
        return {
            "rows_before": len(df_before),
            "rows_after": len(df_after),
            "rows_removed": len(df_before) - len(df_after),
            "cols_before": len(df_before.columns),
            "cols_after": len(df_after.columns),
            "cols_removed": len(df_before.columns) - len(df_after.columns)
        }