读取全部数据，显示前5行
import pandas as pddef csv_to_df(file):pd.set_option('display.max_columns', None) # 显示所有列pd.set_option('display.max_rows', None) # 显示所有行pd.set_option('display.width', None) # 显示宽度是无限wine_df = pd.read_csv(file)print(wine_df.head())if __name__ == '__main__':filename = '../data/winemag-data-130k-v2.csv'csv_to_df(filename)
number country description designation points price province region_1 region_2 taster_name taster_twitter_handle title variety winery0 0 Italy Aromas include tropical fruit, broom, brimston... Vulkà Bianco 87 NaN Sicily & Sardinia Etna NaN Kerin O’Keefe @kerinokeefe Nicosia 2013 Vulkà Bianco (Etna) White Blend Nicosia1 1 Portugal This is ripe and fruity, a wine that is smooth... Avidagos 87 15.0 Douro NaN NaN Roger Voss @vossroger Quinta dos Avidagos 2011 Avidagos Red (Douro) Portuguese Red Quinta dos Avidagos2 2 US Tart and snappy, the flavors of lime flesh and... NaN 87 14.0 Oregon Willamette Valley Willamette Valley Paul Gregutt @paulgwine Rainstorm 2013 Pinot Gris (Willamette Valley) Pinot Gris Rainstorm3 3 US Pineapple rind, lemon pith and orange blossom ... Reserve Late Harvest 87 13.0 Michigan Lake Michigan Shore NaN Alexander Peartree NaN St. Julian 2013 Reserve Late Harvest Riesling ... Riesling St. Julian4 4 US Much like the regular bottling from 2012, this... Vintner's Reserve Wild Child Block 87 65.0 Oregon Willamette Valley Willamette Valley Paul Gregutt @paulgwine Sweet Cheeks 2012 Vintner's Reserve Wild Child... Pinot Noir Sweet Cheeks
自定义显示的列名
import pandas as pddef csv_to_df(file):pd.set_option('display.max_columns', None) # 显示所有列pd.set_option('display.max_rows', None) # 显示所有行pd.set_option('display.width', None) # 显示宽度是无限wine_df = pd.read_csv(file, skiprows=1,names=['编号','国家','描述','标识','评分','价格','省份','地区1','地区2','品酒师','twitter帐号','名称','品种','酒厂'])print(wine_df.head())if __name__ == '__main__':filename = '../data/winemag-test.csv'csv_to_df(filename)
编号 国家 描述 标识 评分 价格 省份 地区1 地区2 品酒师 twitter帐号 名称 品种 酒厂0 0 Italy Aromas include tropical fruit, broom, brimston... Vulkà Bianco 87.0 NaN Sicily & Sardinia Etna NaN Kerin O’Keefe @kerinokeefe Nicosia 2013 Vulkà Bianco (Etna) White Blend Nicosia1 1 Portugal This is ripe and fruity, a wine that is smooth... Avidagos 87.0 15.0 Douro NaN NaN Roger Voss @vossroger Quinta dos Avidagos 2011 Avidagos Red (Douro) Portuguese Red Quinta dos Avidagos2 2 US Tart and snappy, the flavors of lime flesh and... NaN 87.0 14.0 Oregon Willamette Valley Willamette Valley Paul Gregutt @paulgwine Rainstorm 2013 Pinot Gris (Willamette Valley) Pinot Gris Rainstorm3 3 NaN Pineapple rind, lemon pith and orange blossom ... Reserve Late Harvest NaN 13.0 Michigan Lake Michigan Shore NaN Alexander Peartree NaN St. Julian 2013 Reserve Late Harvest Riesling ... Riesling St. Julian4 4 US Much like the regular bottling from 2012, this... Vintner's Reserve Wild Child Block 87.0 65.0 Oregon Willamette Valley Willamette Valley Paul Gregutt @paulgwine Sweet Cheeks 2012 Vintner's Reserve Wild Child... Pinot Noir Sweet Cheeks
选择输出指定列的数据
import pandas as pddef csv_to_df(file):pd.set_option('display.max_columns', None) # 显示所有列pd.set_option('display.max_rows', None) # 显示所有行pd.set_option('display.width', None) # 显示宽度是无限wine_df = pd.read_csv(file, usecols=['number','country','points','price','variety'])print(wine_df.head(10))if __name__ == '__main__':filename = '../data/winemag-test.csv'csv_to_df(filename)
number country points price variety0 0 Italy 87.0 NaN White Blend1 1 Portugal 87.0 15.0 Portuguese Red2 2 US 87.0 14.0 Pinot Gris3 3 NaN NaN 13.0 Riesling4 4 US 87.0 65.0 Pinot Noir5 5 Spain 87.0 15.0 NaN6 6 Italy 87.0 16.0 Frappato7 7 France 87.0 24.0 Gewürztraminer8 8 Germany 87.0 12.0 Gewürztraminer9 9 France 87.0 27.0 Pinot Gris
查看缺失数据情况
import pandas as pddef csv_to_df(file):wine_df = pd.read_csv(file, usecols=['number','country','points','price','variety'])print(wine_df.head(10))print(wine_df.country[3])print(wine_df.points[3])print(wine_df.iloc[3, 2])print(type(wine_df.iloc[3,2]))print(type(wine_df.points[3]))if __name__ == '__main__':filename = '../data/winemag-test.csv'csv_to_df(filename)
number country points price variety0 0 Italy 87.0 NaN White Blend1 1 Portugal 87.0 15.0 Portuguese Red2 2 US 87.0 14.0 Pinot Gris3 3 NaN NaN 13.0 Riesling4 4 US 87.0 65.0 Pinot Noir5 5 Spain 87.0 15.0 NaN6 6 Italy 87.0 16.0 Frappato7 7 France 87.0 24.0 Gewürztraminer8 8 Germany 87.0 12.0 Gewürztraminer9 9 France 87.0 27.0 Pinot Grisnannannan<class 'numpy.float64'><class 'numpy.float64'>
填充缺失值
DataFrame.fillna(value=None, method=None, axis=None, inplace=False, limit=None, downcast=None)
inplace=False：默认值，填充后产生新的dataframe，用的时候需要重新命名
inplace=True: 原地填充，直接改变当前dataframe的数据
value：scalar, dict, Series, or DataFrame
用来填充缺失的值（例如0），可以是dict、Series或DataFrame，指定每个索引（对于序列）或列（对于DataFrame）使用哪个值。不在dict、Series或DataFrame中的值将不会被填充。此值不能是列表。
method：{‘backfill’, ‘bfill’, ‘pad’, ‘ffill’, None}, default None
用于在重新编制索引的系列垫中填充缺失值的方法
pad / ffill表示用前面行/列的值，填充当前行/列的空值，
backfill / bfill表示用后面行/列的值，填充当前行/列的空值。
这里的前、后一个数值默认是纵向看的，如果需要使用左或者右边的数值进行填充，只需要加参数axis=1，就可以了。
axis：{0 or ‘index’, 1 or ‘columns’}
确定填充缺失值时的轴的方向，0 或‘index’为纵向, 1或 ‘columns’为横向填充。
inplace：bool, 缺省 False
缺省情况下不改变当前dataframe对象，创建 一个新的对象并填充，如果设为True，直接在当前dataframe中填充数据，值为True时返回值是None，值为False时返回值是填充后新的dataframe.
limit：int, 缺省 None
用limit限制每列可以替代NaN的数目，下面我们限制每列只能替代一个NaN。如果method被指定，对于连续的空值，这段连续区域，最多填充前 limit 个空值（如果存在多段连续区域，每段最多填充前 limit 个空值）。如果method未被指定， 在该axis下，最多填充前 limit 个空值（不论空值连续区间是否间断）
downcast：dict, 缺省 None
字典中的项为，为类型向下转换规则。或者为字符串“infer”，此时会在合适的等价类型之间进行向下转换，比如float64转 int64 。
import pandas as pddef csv_to_df(file):wine_df = pd.read_csv(file, usecols=['number','country','points','price','variety'])wine_df.fillna(0, inplace=True)print(wine_df.head(10))if __name__ == '__main__':filename = '../data/winemag-test.csv'csv_to_df(filename)
number country points price variety0 0 Italy 87.0 0.0 White Blend1 1 Portugal 87.0 15.0 Portuguese Red2 2 US 87.0 14.0 Pinot Gris3 3 0 0.0 13.0 Riesling4 4 US 87.0 65.0 Pinot Noir5 5 Spain 87.0 15.0 06 6 Italy 87.0 16.0 Frappato7 7 France 87.0 24.0 Gewürztraminer8 8 Germany 87.0 12.0 Gewürztraminer9 9 France 87.0 27.0 Pinot Gris
fillna可以指定数值进行填充，也可以使用计算公式进行填充，比如df.mean()、df.sum()等。
还可以指定用那一列的数据进行填充，例如只填充price列：
import pandas as pddef csv_to_df(file):wine_df = pd.read_csv(file, usecols=['number','country','points','price','variety'])wine_df.price.fillna(0, inplace=True)print(wine_df.head(10))if __name__ == '__main__':filename = '../data/winemag-test.csv'csv_to_df(filename)
number country points price variety0 0 Italy 87.0 0.0 White Blend1 1 Portugal 87.0 15.0 Portuguese Red2 2 US 87.0 14.0 Pinot Gris3 3 NaN NaN 13.0 Riesling4 4 US 87.0 65.0 Pinot Noir5 5 Spain 87.0 15.0 NaN6 6 Italy 87.0 16.0 Frappato7 7 France 87.0 24.0 Gewürztraminer8 8 Germany 87.0 12.0 Gewürztraminer9 9 France 87.0 27.0 Pinot Gris
用价格的平均值填充价格列：
import pandas as pddef csv_to_df(file):wine_df = pd.read_csv(file, usecols=['number','country','points','price','variety'])wine_df.price.fillna(wine_df.price.mean(), inplace=True)print(wine_df.head(10))if __name__ == '__main__':filename = '../data/winemag-test.csv'csv_to_df(filename)
number country points price variety0 0 Italy 87.0 35.363389 White Blend1 1 Portugal 87.0 15.000000 Portuguese Red2 2 US 87.0 14.000000 Pinot Gris3 3 NaN NaN 13.000000 Riesling4 4 US 87.0 65.000000 Pinot Noir5 5 Spain 87.0 15.000000 NaN6 6 Italy 87.0 16.000000 Frappato7 7 France 87.0 24.000000 Gewürztraminer8 8 Germany 87.0 12.000000 Gewürztraminer9 9 France 87.0 27.000000 Pinot Gris
用后一个值填充
import pandas as pddef csv_to_df(file):wine_df = pd.read_csv(file, usecols=['number','country','points','price','variety'])wine_df.points.fillna(method='bfill', inplace=True)print(wine_df.head(10))if __name__ == '__main__':filename = '../data/winemag-test.csv'csv_to_df(filename)
number country points price variety0 0 Italy 83.0 NaN White Blend1 1 Portugal 87.0 15.0 Portuguese Red2 2 US 89.0 14.0 Pinot Gris3 3 NaN 85.0 13.0 Riesling4 4 US 85.0 65.0 Pinot Noir5 5 Spain 87.0 15.0 NaN6 6 Italy 97.0 16.0 Frappato7 7 France 87.0 24.0 Gewürztraminer8 8 Germany 80.0 12.0 Gewürztraminer9 9 France 87.0 27.0 Pinot Gris
用前一个值进行填充
import pandas as pddef csv_to_df(file):wine_df = pd.read_csv(file, usecols=['number','country','points','price','variety'])wine_df.points.fillna(method='ffill', inplace=True)print(wine_df.head(10))if __name__ == '__main__':filename = '../data/winemag-test.csv'csv_to_df(filename)
number country points price variety0 0 Italy 83.0 NaN White Blend1 1 Portugal 87.0 15.0 Portuguese Red2 2 US 89.0 14.0 Pinot Gris3 3 NaN 89.0 13.0 Riesling4 4 US 85.0 65.0 Pinot Noir5 5 Spain 87.0 15.0 NaN6 6 Italy 97.0 16.0 Frappato7 7 France 87.0 24.0 Gewürztraminer8 8 Germany 80.0 12.0 Gewürztraminer9 9 France 87.0 27.0 Pinot Gris
若有收获，就点个赞吧
0 人点赞
