Python如何修复Pandas groupby对象上的“ValueError:尝试获取空序列的argmin”

Question

''' 我正试图基于df.groupby（'Chrom'）对象上的'Position'列创建容器。 我在代码的最后一部分上苦苦挣扎了几个小时。脚本的目的是为groupby对象的每个组检索最低的pvalue（p）。 df_lowestP=data.loc[data.groupby（'Bin_labels'）['p'].idxmin（）]引发了以下错误“ValueError:尝试获取空序列的argmin” 我不知道如何识别空序列以及如何进行排序。 提前谢谢你帮了我的忙

'''

cols= ['Chrom', 'Position', 'Marker', 'p']
vals = [['5A', 70.86, 'wsnp_Ex_c2474_4619730', 0.952967047],\
['2D', 80.46, 'Excalibur_c48871_1330', 0.603140475],\
['5B', 21.6, 'BS00010491_51', 0.68806773],\
['1D', 7.47, 'BobWhite_c4303_524', 0.14664623599999999],\
['6B', 52.36, 'RAC875_c54904_460', 0.747116322],\
['1B', 56.92, 'BS00023044_51', 0.793591901],\
['1B', 82.21, 'Excalibur_c60931_1260', 0.10407994],\
['7B', 61.76, 'Kukri_c12822_132', 0.9914339959999999],\
['6B', 99.03, 'BS00069285_51', 0.591267434],\
['6B', 65.68, 'Tdurum_contig42414_612', 0.6802424490000001],\
['2A', 0.0, 'wsnp_Ex_c1962_3696626', 0.016681088],\
['2D', 88.72, 'BS00031459_51', 0.06155213599999999],\
['3B', 60.02, 'RAC875_c2787_470', 0.359986433],\
['2B', 89.94, 'Excalibur_c16646_1254', 0.548888715],\
['6D', 48.85, 'Excalibur_c8134_1319', 0.4144683],\
['5B', 9.03, 'wsnp_Ex_c2459_4591587', 0.64927464],\
['5B', 123.43, 'BS00021868_51', 0.478334741],\
['2A', 194.55, 'wsnp_CAP11_rep_c6137_2841945', 0.67702519],\
['2B', 97.48, 'RAC875_c46788_60', 0.461550513],\
['1A', 73.26, 'CAP7_c3269_236', 0.8397169459999999],\
['1B', 148.6, 'RAC875_c9082_267', 0.992387433],\
['3A', 158.94, 'RAC875_c3084_415', 0.35521176200000004],\
['2A', 77.73, 'Kukri_c7139_1248', 0.865599771],\
['4A', 55.63, 'wsnp_Ex_c829_1620518', 0.628465917],\
['4A', 61.94, 'Excalibur_c96303_224', 0.8707948609999999],\
['1A', 85.79, 'Excalibur_c13183_128', 0.811616063],\
['3A', 79.22, 'wsnp_Ex_c32003_40728918', 0.45561401100000004],\
['7B', 29.87, 'Excalibur_c54242_704', 0.165007274],\
['6B', 75.73, 'Kukri_c60169_54', 0.893133473],\
['7B', 94.94, 'BS00065981_51', 0.45571674],\
['6A', 64.26, 'GENE-3933_586', 0.974289783],\
['1B', 0.0, 'Kukri_c44369_131', 0.311861047],\
['5D', 9.36, 'IACX2960', 0.9434061220000001],\
['6D', 48.34, 'Ex_c12134_939', 0.456392529],\
['3B', 97.16, 'wsnp_Ex_c18915_27811736', 0.785250982],\
['5B', 166.0, 'BS00098520_51', 0.6439036220000001],\
['2B', 96.44, 'RAC875_rep_c111100_447', 0.063204845],\
['1A', 51.88, 'Kukri_c3150_341', 0.17333784600000002],\
['5A', 139.08, 'BS00010179_51', 0.8453566920000001],\
['1B', 126.13, 'wsnp_Ku_c18881_28259811', 0.653103781],\
['1B', 128.64, 'Excalibur_c8585_701', 0.103269216],\
['2D', 44.17, 'Excalibur_c40820_335', 0.08864491],\
['1B', 80.3, 'Kukri_c5335_2165', 0.567401681],\
['3A', 97.15, 'RAC875_c62012_382', 0.407825859],\
['4B', 52.36, 'GENE-0037_298', 0.014111186999999999],\
['5A', 79.55, 'Tdurum_contig59930_366', 0.117716542],\
['5B', 128.25, 'BS00028082_51', 0.24427141],\
['7B', 109.87, 'wsnp_Ku_rep_c103690_90365438', 0.670004615],\
['2D', 89.68, 'BS00066526_51', 0.025921416],\
['7A', 81.25, 'IAAV5328', 0.340900239]]

df = pd.DataFrame(vals, columns=cols)

for gp,data in df.groupby('Chrom'):
    pos_range = data['Position'].max() - data['Position'].min()
    inter_value = 5 
    min_value = int(np.floor(data['Position'].min()))
    max_value = int(np.ceil(data['Position'].max()))
    intervals = [i for i in range(min_value, max_value+inter_value, inter_value)] #creating bins boundaries
    labels = ['Bin_' + str(i) for i in range(1, len(intervals))] #labelling bins
    #creating bin_labels column
    bin_labels = pd.cut(x=data['Position'],
                                    bins=intervals,
                                    labels=labels,
                                    include_lowest=True)
    # creating bin boundaries column
    bin_range = pd.cut(x=data['Position'],
                             bins=intervals,
                             include_lowest=True)

    data.insert(3,'Bin_labels',bin_labels)
    data.insert(4,'Bin_range',bin_range)
    
    #Retrieving a single marker with lowest P for each bin
    df_lowestP = data.loc[data.groupby('Bin_labels')['p'].idxmin()] #Here is the trouble

Answer 1

您的Bin_labels是分类类型。它包括值为NaN的其他箱子。因此，您得到了空序列。当您按Bin_labels分组时，会丢失原始索引

要解决这个问题，我们必须保存原始索引，按Bin_labels和p对数据进行排序，按Bin_labels分组，并获取第一个条目。不确定这是否是最好的解决方案

data['idx'] = data.index
data = data.sort_values(['Bin_labels', 'p'])
idx = data.groupby('Bin_labels').first()['idx'].dropna()
df_lowestP = data.loc[idx]

这是第二组

原件：

Name: idx, dtype: float64
   Chrom  Position                   Marker Bin_labels       Bin_range         p  idx
31    1B      0.00         Kukri_c44369_131      Bin_1   (-0.001, 5.0]  0.311861   31
5     1B     56.92            BS00023044_51     Bin_12    (55.0, 60.0]  0.793592    5
6     1B     82.21    Excalibur_c60931_1260     Bin_17    (80.0, 85.0]  0.104080    6
42    1B     80.30         Kukri_c5335_2165     Bin_17    (80.0, 85.0]  0.567402   42
40    1B    128.64      Excalibur_c8585_701     Bin_26  (125.0, 130.0]  0.103269   40
39    1B    126.13  wsnp_Ku_c18881_28259811     Bin_26  (125.0, 130.0]  0.653104   39
20    1B    148.60         RAC875_c9082_267     Bin_30  (145.0, 150.0]  0.992387   20

最低：

     Chrom  Position                 Marker Bin_labels       Bin_range         p  idx
31.0    1B      0.00       Kukri_c44369_131      Bin_1   (-0.001, 5.0]  0.311861   31
5.0     1B     56.92          BS00023044_51     Bin_12    (55.0, 60.0]  0.793592    5
6.0     1B     82.21  Excalibur_c60931_1260     Bin_17    (80.0, 85.0]  0.104080    6
40.0    1B    128.64    Excalibur_c8585_701     Bin_26  (125.0, 130.0]  0.103269   40
20.0    1B    148.60       RAC875_c9082_267     Bin_30  (145.0, 150.0]  0.992387   20