我想根据年份拆分我的数据框

Question

我有一个包含 datetime64 格式的日期值列的数据框。我想根据年份将我的数据帧拆分为单独的数据帧。我写了下面的代码，它有效但非常不切实际。希望有人有更好的解决方案！# import libsimport numpy as npimport pandas as pdfrom random import sample# Make some random dataframe with two columnsdate = np.arange('2005-02', '2008-03', dtype='datetime64[D]')status = ["X"]*(int(round(0.9*len(date),0))) +['y']*(int(round(0.05*len(date),0)))+['z']*(int(round(0.05*len(date),0)))newstatus = sample(status, len(status))data = {'Data': date, 'Status': newstatus}df = pd.DataFrame(data)# Extract year from date and make dummies index for splittingdf['Year'] = pd.DatetimeIndex(df['Data']).yeardf = pd.get_dummies(df, columns = ['Year'])# Split on dummiesdf_2007, df_2006, df_2005, df_2008  = df, df, df, dfdf_2008= df_2008[df_2008.Year_2008 != 0]df_2007 = df_2007[df_2007.Year_2007 != 0]df_2006= df_2006[df_2006.Year_2006 != 0]df_2005= df_2005[df_2005.Year_2005 != 0]#Remove Dummiesyears = ['Year_2005', 'Year_2006', 'Year_2007', 'Year_2008']df_2008 = df_2008.drop(years, axis = 1)df_2007 = df_2007.drop(years, axis = 1)df_2006 = df_2006.drop(years, axis = 1)df_2005 = df_2005.drop(years, axis = 1)

Answer 1

也许这可以帮助你：

years = df['Data'].dt.year.unique() # I'm guessing Data should be Date really but I'll go along with it.

dfs = {y: df[df['Data'].dt.year == y] for y in years}

这将创建一个字典，其中键是年份，值是对应于每一年的数据框。这意味着dfs[2008]为您提供包含 2008 年数据的数据框。