python的dataframe常用处理方法

import pandas as pdclass DataFrameProcessor:@staticmethoddef sort_by_column(df, by_column, ascending=True):"""根据指定列对DataFrame进行排序。Parameters:df (pd.DataFrame): 要排序的DataFrame。by_column (str): 要排序的列名。ascending (bool): True表示升序排列，False表示降序排列。Returns:pd.DataFrame: 排序后的新DataFrame。"""sorted_df = df.sort_values(by=by_column, ascending=ascending)return sorted_df@staticmethoddef remove_rows_with_value(df, column_name, value_to_remove):"""从DataFrame中删除指定列等于特定值的所有行。Parameters:df (pd.DataFrame): 要处理的DataFrame。column_name (str): 要检查的列名。value_to_remove: 要删除的特定值。Returns:pd.DataFrame: 新的DataFrame，不包含指定列等于特定值的行。"""filtered_df = df[df[column_name] != value_to_remove]return filtered_df@staticmethoddef drop_duplicates(df, subset=None):"""去除DataFrame中的重复行。Parameters:df (pd.DataFrame): 要处理的DataFrame。subset (list): 可选，要考虑的列的子集。Returns:pd.DataFrame: 不包含重复行的新DataFrame。"""deduplicated_df = df.drop_duplicates(subset=subset)return deduplicated_df@staticmethoddef get_stats(df):"""计算DataFrame的统计信息。Parameters:df (pd.DataFrame): 要计算统计信息的DataFrame。Returns:pd.Series: 包含统计信息的Series。"""return df.describe()@staticmethoddef get_missing(df):"""计算DataFrame中每列的缺失值数量。Parameters:df (pd.DataFrame): 要计算缺失值的DataFrame。Returns:pd.Series: 包含每列缺失值数量的Series。"""return df.isnull().sum()@staticmethoddef fill_missing(df, value):"""填充DataFrame中的缺失值。Parameters:df (pd.DataFrame): 要处理的DataFrame。value: 用于填充缺失值的值。Returns:pd.DataFrame: 填充缺失值后的新DataFrame。"""filled_df = df.fillna(value)return filled_df@staticmethoddef filter(df, query):"""根据条件过滤DataFrame。Parameters:df (pd.DataFrame): 要过滤的DataFrame。query (str): 过滤条件。Returns:pd.DataFrame: 符合条件的新DataFrame。"""filtered_df = df.query(query)return filtered_df@staticmethoddef convert_data_types(df, conversion_dict):"""将DataFrame的列的数据类型进行转换。Parameters:df (pd.DataFrame): 要处理的DataFrame。conversion_dict (dict): 包含列名和目标数据类型的字典。Returns:pd.DataFrame: 数据类型转换后的新DataFrame。"""converted_df = df.astype(conversion_dict)return converted_df@staticmethoddef rename_columns(df, column_mapping):"""重命名DataFrame的列。Parameters:df (pd.DataFrame): 要处理的DataFrame。column_mapping (dict): 包含原列名和新列名的字典。Returns:pd.DataFrame: 列重命名后的新DataFrame。"""renamed_df = df.rename(columns=column_mapping)return renamed_df@staticmethoddef calculate_mean_and_std(df, column_name):"""计算DataFrame中指定列的均值和3倍标准差。Parameters:df (pd.DataFrame): 要处理的DataFrame。column_name (str): 要计算统计值的列名。Returns:tuple: 包含均值和3倍标准差的元组 (mean, std)."""column_data = df[column_name]mean_value = column_data.mean()std_value = column_data.std() * 3  # 3倍标准差return mean_value, std_value# 示例用法：# 创建一个示例DataFrame
data = {'A': [1, 4, 2, 5, 3],'B': ['apple', 'cherry', 'banana', 'banana', 'apple']}df = pd.DataFrame(data)# 使用DataFrameProcessor类的方法
sorted_df = DataFrameProcessor.sort_by_column(df, 'A', ascending=True)
filtered_df = DataFrameProcessor.remove_rows_with_value(df, 'B', 'apple')
deduplicated_df = DataFrameProcessor.drop_duplicates(df)
stats = DataFrameProcessor.get_stats(df)
missing = DataFrameProcessor.get_missing(df)
filled_df = DataFrameProcessor.fill_missing(df, value=0)
filtered_df = DataFrameProcessor.filter(df, query='A > 2')
converted_df = DataFrameProcessor.convert_data_types(df, {'A': float})
renamed_df = DataFrameProcessor.rename_columns(df, {'A': 'New_A'})
mean, std = DataFrameProcessor.calculate_mean_and_std(df, 'A')print(sorted_df)
print(filtered_df)
print(deduplicated_df)
print(stats)
print(missing)
print(filled_df)
print(filtered_df)
print(converted_df)
print(renamed_df)
print(f"均值: {mean}, 3倍标准差: {std}")

这个完整的 DataFrameProcessor 类包含了各种常见的数据处理方法，可以根据需要使用。