下面是我的数据框：Df1
origin 2001-01-01 00:00:00 2002-01-01 00:00:00 2003-01-01 00:00:00 ... 2009-01-01 00:00:00 2010-01-01 00:00:00 Grand Total
Simulation 1 2.281294e+13 NaN 1.459444e+20 ... 1.618202e+59 6.811895e+64 1.748673e+72
Simulation 2 2.175246e+13 NaN 1.711078e+20 ... 1.388109e+59 5.608744e+64 2.106283e+72
Simulation 3 2.280800e+13 NaN 1.422197e+20 ... 9.737150e+58 2.589176e+64 9.651312e+71
Simulation 4 2.426074e+13 NaN 1.369407e+20 ... 1.063072e+59 2.921763e+64 1.961873e+71
第二个数据框：df2
CI Var
0 60.0 3.009765e+72
1 70.0 3.511393e+72
2 80.0 4.013020e+72
3 90.5 4.539729e+72
下面是我的代码：
class cl_test():
@classmethod
def resamledtriangle(self,n_sims):
# Load triangle
triangle = cl.load_sample('genins')
# Use bootstrap sampler to get resampled triangles
bootstrapdataframe = cl.BootstrapODPSample(n_sims = n_sims, random_state=42).fit(triangle).resampled_triangles_
#converting to dataframe
resampledtriangledf1 = bootstrapdataframe.to_frame()
#print(resampledtriangledf1)
Reserve_val = []
iter_count=0
while (iter_count < n_sims):
bg = resampledtriangledf1[resampledtriangledf1.index==iter_count]
resampledtriangledf = bg.pivot_table(index="origin", columns="development", values="values")
#print(resamp)
#final = resampledtriangledf[::-1].bfill().iloc[0]
final2 = resampledtriangledf.ffill(axis=1).iloc[:, -1]
resampledtriangledf = resampledtriangledf.assign(Total = final2.values)
#resampledtriangledf = resampledtriangledf.append(pd.Series(totTriangleDf1.sum(), name='Total')).append(pd.Series(final, name='Latest Observation'))
#calculating grand total of each row and column
#resampledtriangledf['Grand Total'] = resampledtriangledf.sum(axis=1)
#resampledtriangledf.loc["Grand Total"] = resampledtriangledf.sum()
resampledtriangledf = resampledtriangledf.append(pd.Series(resampledtriangledf.sum(), name='Grand Total'))
resampledtriangcdfwa1 = resampledtriangledf.shift(-1, axis=1).loc['Grand Total'] / (resampledtriangledf.loc['Grand Total'])
resampledtriangledf = resampledtriangledf.append(pd.Series(resampledtriangcdfwa1,name='Weighted Average'))
c = resampledtriangledf.iloc[1][::-1].replace(np.inf, np.nan).dropna().cumprod()
resampledtriangledf = resampledtriangledf.append(c.rename('CDF'))
s = resampledtriangledf.iloc[:, :][:-2].copy()
ultiCalc = pd.DataFrame()
ultiCalc['Latest']= s['Total']
ultiCalc['CDF'] = np.flip(resampledtriangledf.loc['CDF'].values)
ultiCalc['Ultimate'] = ultiCalc['Latest']*ultiCalc['CDF']
ultiCalc['Unpaid'] = ultiCalc['Ultimate']-ultiCalc['Latest']
ultiCalc = ultiCalc.append(pd.Series(ultiCalc.sum(), name='Total'))
#ultiCalc = ultiCalc.iloc[: , :-1]
iter_count+=1
Reserve_val.append(ultiCalc['Unpaid'])
simulationDf = pd.DataFrame([x for x in Reserve_val]).reset_index(drop=True)
simulationDf.index = 'Simulation ' + (simulationDf.index + 1).astype(str)
simulationDf = simulationDf.iloc[: , :-1]
return simulationDf
def varmethod(self,CI,simulationDf):
total = simulationDf['Grand Total'].sum()
df = pd.DataFrame(CI,columns=['CI'])
for value,data in enumerate(CI): #== The enumerate() method adds counter to an iterable and returns it (the enumerate object).
CI[value]=(data/100)*(total)
df['Var'] = CI
print(df)
return df
在这里，我必须从 df2 迭代 var 值并与 df 数据框的总计列进行比较。并创建新的数据框并存储所有大于 Var 值的总计值。例如：Var 列的第一个值为 3.009765e+72。我必须从 df1 create dataframe 的总计中过滤所有更大的值。接下来我们必须过滤 Var(3.511393e+72) 的第二个值并创建数据帧。很快