熊貓追加過濾行到另一個數據幀

Question

我有2個熊貓數據幀df和df_min。我將一些過濾器應用於df，這會產生一行數據，並且我想將該行附加到df_min。我嘗試使用循環遍歷df，並嘗試使用loc將該行追加到df_min。我一直得到Incompatible indexer with DataFrame ValueError的行，我使用loc。我想我沒有正確使用loc。什麼是最好的方式來完成我想要做的事情？

i = 0 
for elem in vehicles: 
    for state in limit_states: 
     a = df[(df.VEHICLE == elem) & (df.LIMIT_STATE == state)] 
     df_min.loc[i] = a[(a.RF == np.min(a.RF))].head(1) #results in a single row 
     i = i + 1 

編輯：我也試過以下，而不是loc，但得到了同樣的錯誤：

df_min.ix[i] = a[(a.RF == np.min(a.RF))].head(1) 

編輯2：試過以下，患上了「第一個參數必須是一個列表類似的熊貓對象，這次你傳遞了一個類型爲「DataFrame」的對象。

for elem in vehicles: 
    for state in limit_states: 
     a = df[(df.VEHICLE == elem) & (df.LIMIT_STATE == state)] 
     df_min = pd.concat(a[(a.RF == np.min(a.RF))].head(1)) 

Answer 1

大概就像這將是有益的：

df_min = pd.concat([ df[(df.VEHICLE == elem) & (df.LIMIT_STATE == state)] 
        for elem in vehicles for state in limit_states ]) 

編輯：根據名單vehicles

xs = [ df[(df.VEHICLE == elem) & (df.LIMIT_STATE == state)] 
        for elem in vehicles for state in limit_states ] 
df_min = pd.concat([ a[(a.RF == np.min(a.RF))].head(1) for a in xs ]) 

和limit_states你大概也能達到什麼你正在嘗試做的使用groupby;例如：

fn = lambda a: a[(a.RF == np.min(a.RF))].head(1) 
df.groupby(['VEHICLE', 'LIMIT_STATE']).apply(fn)