是什么导致这个 KeyError 来自一个简单的函数调用？答案

【问题标题】：What is causing this KeyError from a simple function call?是什么导致这个 KeyError 来自一个简单的函数调用？
【发布时间】：2018-12-27 02:19:04
【问题描述】：

以下错误是由以下代码引起的。我读过KeyError: 0 is due to a dictionary file lacking an entry，但我仍然不知道字典文件是什么或我的代码如何访问它：我只是想访问数据框中的数据。显然问题是数据帧的一个子集VolValues 使用从 23000 左右开始的索引，而我试图用索引 '0' 对它进行切片，因为我认为这是 python 的“第一个元素”语法。

你能告诉我代码有什么问题以及如何修复它吗？

runfile('/Users/daniel/Documents/programming/RectumD2Metrics.py', wdir='/Users/daniel/Documents/programming')
Traceback (most recent call last):

  File "<ipython-input-2-d170dca123d7>", line 1, in <module>
    runfile('/Users/daniel/Documents/programming/RectumD2Metrics.py', wdir='/Users/daniel/Documents/programming')

  File "/anaconda3/lib/python3.6/site-packages/spyder/utils/site/sitecustomize.py", line 705, in runfile
    execfile(filename, namespace)

  File "/anaconda3/lib/python3.6/site-packages/spyder/utils/site/sitecustomize.py", line 102, in execfile
    exec(compile(f.read(), filename, 'exec'), namespace)

  File "/Users/daniel/Documents/programming/RectumD2Metrics.py", line 37, in <module>
    D2Planned = interpD2('planned',df)

  File "/Users/daniel/Documents/programming/RectumD2Metrics.py", line 30, in interpD2
    if (VolValues[loop] > 2) and (VolValues[loop+1] < 2):

  File "/anaconda3/lib/python3.6/site-packages/pandas/core/series.py", line 766, in __getitem__
    result = self.index.get_value(self, key)

  File "/anaconda3/lib/python3.6/site-packages/pandas/core/indexes/base.py", line 3103, in get_value
    tz=getattr(series.dtype, 'tz', None))

  File "pandas/_libs/index.pyx", line 106, in pandas._libs.index.IndexEngine.get_value

  File "pandas/_libs/index.pyx", line 114, in pandas._libs.index.IndexEngine.get_value

  File "pandas/_libs/index.pyx", line 162, in pandas._libs.index.IndexEngine.get_loc

  File "pandas/_libs/hashtable_class_helper.pxi", line 958, in pandas._libs.hashtable.Int64HashTable.get_item

  File "pandas/_libs/hashtable_class_helper.pxi", line 964, in pandas._libs.hashtable.Int64HashTable.get_item

KeyError: 0

代码：

# Import Rectum DVH
import pandas
import numpy
df = pandas.read_csv('/Users/daniel/Documents/data/DVH/RectumData.csv',
                 delimiter=',',header=0)
# Calculate D_2%, defined by ICRU 78 as "the greatest dose which all but
# 2 percent of a [volume of interest] receives." aka D_{near-max}

def interpD2(disttype,df):
# Loop through all patients' plans.
    Dose2Results = numpy.zeros(40)
    for num in range(0,40): 
# We know a priori that there is no DVH data with Volume = 2. Hence we look for
# the two columns less than and greater than Volume = 2.
        if disttype == 'planned':
            DoseValues = df.loc[(df['StudyID'] == num+1) & (df['DistributionType'] == 'planned')].Dose
            VolValues = df.loc[(df['StudyID'] == num+1) & (df['DistributionType'] == 'planned')].Volume
        else:
            DoseValues = df.loc[(df['StudyID'] == num+1) & (df['DistributionType'] == 'blurred')].Dose
            VolValues = df.loc[(df['StudyID'] == num+1) & (df['DistributionType'] == 'blurred')].Volume
        for loop in range(0,len(VolValues)):
            if (VolValues[loop] > 2) and (VolValues[loop+1] < 2):
                LowerVolumeIndex,UpperVolumeIndex = loop,loop+1
                x0,x1,x2 = 2,VolValues[LowerVolumeIndex],VolValues[UpperVolumeIndex]
                y1,y2 = DoseValues[LowerVolumeIndex],DoseValues[UpperVolumeIndex]
                Dose2Results[num] = y1 - ((x1-x0)/(x2 - x1))*(y2 - y1)
    return Dose2Results
D2Planned,D2Blurred = numpy.zeros(40),numpy.zeros(40)
D2Planned = interpD2('planned',df)
D2Blurred = interpD2('blurred',df)

导入的 CSV 文件的摘录在这篇文章的末尾。

尝试解决：

删除传递给函数的df 会导致相同的错误消息。（它最初不存在，因为我认为函数可以访问“全局”变量。）
我尝试用零“初始化”变量。
我使用 if 块显式解析了字符串，试图解决错误消息，但仍然没有任何变化。

检查the pandas page，我发现有可用的更新。我已经通过conda install pandas 安装了它，但错误消息仍然没有变化。（此更新的详细信息如下。）

The following packages will be downloaded:

package                    |            build
---------------------------|-----------------
certifi-2018.4.16          |           py36_0         142 KB
conda-4.5.8                |           py36_0         1.0 MB
------------------------------------------------------------
                                       Total:         1.2 MB

The following packages will be UPDATED:

certifi: 2018.4.16-py36_0 conda-forge --> 2018.4.16-py36_0
conda:   4.5.6-py36_0     conda-forge --> 4.5.8-py36_0

感谢您的帮助。

CSV 数据摘录，包括第一个标题行；我已跳过行以保持在正文帖子字符限制内，但请注意首先列出 DistributionType，然后是 StudyID。因此，数字变为 1,10,11,...,19,2,20,21,...，“模糊”数据位于“计划”数据之前。

StudyID,DistributionType,Organ,Dose,Volume,DoseUnit,VolumeUnit
1,blurred,Rectum,0,100,Gy(RBE),%
1,blurred,Rectum,0.1,78.13818,Gy(RBE),%
1,blurred,Rectum,0.2,75.901,Gy(RBE),%
1,blurred,Rectum,0.3,75.01312,Gy(RBE),%
1,blurred,Rectum,0.4,73.38642,Gy(RBE),%
1,blurred,Rectum,0.5,72.36015,Gy(RBE),%
1,blurred,Rectum,0.6,70.81651,Gy(RBE),%
1,blurred,Rectum,7.3,22.60766,Gy(RBE),%
1,blurred,Rectum,7.4,22.4557,Gy(RBE),%
1,blurred,Rectum,7.5,22.31794,Gy(RBE),%
1,blurred,Rectum,7.6,22.19247,Gy(RBE),%
1,blurred,Rectum,7.7,22.09406,Gy(RBE),%
1,blurred,Rectum,32.2,6.99686,Gy(RBE),%
1,blurred,Rectum,32.3,6.96634,Gy(RBE),%
1,blurred,Rectum,32.4,6.94046,Gy(RBE),%
1,blurred,Rectum,32.5,6.89926,Gy(RBE),%
1,blurred,Rectum,32.6,6.85925,Gy(RBE),%
1,blurred,Rectum,32.7,6.83843,Gy(RBE),%
1,blurred,Rectum,32.8,6.8082,Gy(RBE),%
1,blurred,Rectum,32.9,6.76663,Gy(RBE),%
1,blurred,Rectum,33,6.72788,Gy(RBE),%
1,blurred,Rectum,33.1,6.6771,Gy(RBE),%
1,blurred,Rectum,33.2,6.62313,Gy(RBE),%
1,blurred,Rectum,33.3,6.57601,Gy(RBE),%
1,blurred,Rectum,42.5,2.96622,Gy(RBE),%
1,blurred,Rectum,42.6,2.9242,Gy(RBE),%
1,blurred,Rectum,42.7,2.87604,Gy(RBE),%
1,blurred,Rectum,42.8,2.83046,Gy(RBE),%
1,blurred,Rectum,42.9,2.78527,Gy(RBE),%
1,blurred,Rectum,43,2.73564,Gy(RBE),%
1,blurred,Rectum,43.1,2.7077,Gy(RBE),%
1,blurred,Rectum,43.2,2.69686,Gy(RBE),%
1,blurred,Rectum,43.3,2.6505,Gy(RBE),%
1,blurred,Rectum,43.4,2.62119,Gy(RBE),%
1,blurred,Rectum,43.5,2.59528,Gy(RBE),%
1,blurred,Rectum,43.6,2.55359,Gy(RBE),%
1,blurred,Rectum,43.7,2.50786,Gy(RBE),%
1,blurred,Rectum,43.8,2.46692,Gy(RBE),%
1,blurred,Rectum,43.9,2.40788,Gy(RBE),%
1,blurred,Rectum,44,2.37622,Gy(RBE),%
1,blurred,Rectum,44.1,2.34098,Gy(RBE),%
1,blurred,Rectum,44.2,2.30527,Gy(RBE),%
1,blurred,Rectum,44.3,2.26972,Gy(RBE),%
1,blurred,Rectum,44.4,2.2384,Gy(RBE),%
1,blurred,Rectum,44.5,2.20512,Gy(RBE),%
1,blurred,Rectum,44.6,2.14891,Gy(RBE),%
1,blurred,Rectum,44.7,2.12178,Gy(RBE),%
1,blurred,Rectum,44.8,2.06922,Gy(RBE),%
1,blurred,Rectum,44.9,2.02836,Gy(RBE),%
1,blurred,Rectum,45,1.99259,Gy(RBE),%
1,blurred,Rectum,45.1,1.98118,Gy(RBE),%
1,blurred,Rectum,45.2,1.92938,Gy(RBE),%
1,blurred,Rectum,45.3,1.88315,Gy(RBE),%
1,blurred,Rectum,45.4,1.85419,Gy(RBE),%
1,blurred,Rectum,45.5,1.81149,Gy(RBE),%
1,blurred,Rectum,45.6,1.77154,Gy(RBE),%
1,blurred,Rectum,45.7,1.73287,Gy(RBE),%
1,blurred,Rectum,45.8,1.68749,Gy(RBE),%
1,blurred,Rectum,45.9,1.65961,Gy(RBE),%
1,blurred,Rectum,46,1.62265,Gy(RBE),%
1,blurred,Rectum,46.1,1.61065,Gy(RBE),%
1,blurred,Rectum,46.2,1.56712,Gy(RBE),%
1,blurred,Rectum,46.3,1.50282,Gy(RBE),%
1,blurred,Rectum,46.4,1.45122,Gy(RBE),%
1,blurred,Rectum,46.5,1.42696,Gy(RBE),%
1,blurred,Rectum,46.6,1.38877,Gy(RBE),%
1,blurred,Rectum,46.7,1.35886,Gy(RBE),%
1,blurred,Rectum,46.8,1.34022,Gy(RBE),%
1,blurred,Rectum,46.9,1.29308,Gy(RBE),%
1,blurred,Rectum,56.5,NaN,Gy(RBE),%
1,blurred,Rectum,56.6,NaN,Gy(RBE),%
1,blurred,Rectum,56.7,NaN,Gy(RBE),%
1,blurred,Rectum,56.8,NaN,Gy(RBE),%
1,blurred,Rectum,56.9,NaN,Gy(RBE),%
1,blurred,Rectum,57,NaN,Gy(RBE),%
1,blurred,Rectum,57.1,NaN,Gy(RBE),%
1,blurred,Rectum,57.2,NaN,Gy(RBE),%
1,blurred,Rectum,57.3,NaN,Gy(RBE),%
1,blurred,Rectum,57.4,NaN,Gy(RBE),%
1,blurred,Rectum,57.5,NaN,Gy(RBE),%
1,blurred,Rectum,57.6,NaN,Gy(RBE),%
1,blurred,Rectum,57.7,NaN,Gy(RBE),%
1,blurred,Rectum,57.8,NaN,Gy(RBE),%
1,blurred,Rectum,57.9,NaN,Gy(RBE),%
1,blurred,Rectum,58,NaN,Gy(RBE),%
1,blurred,Rectum,58.1,NaN,Gy(RBE),%
1,blurred,Rectum,58.2,NaN,Gy(RBE),%
9,blurred,Rectum,58.2,NaN,Gy(RBE),%
1,planned,Rectum,0,100,Gy(RBE),%
1,planned,Rectum,0.1,78.01999,Gy(RBE),%
1,planned,Rectum,0.2,76.2245,Gy(RBE),%
1,planned,Rectum,14,19.50103,Gy(RBE),%
1,planned,Rectum,14.1,19.4464,Gy(RBE),%
1,planned,Rectum,14.2,19.39261,Gy(RBE),%
1,planned,Rectum,14.3,19.32695,Gy(RBE),%
1,planned,Rectum,14.4,19.25388,Gy(RBE),%
1,planned,Rectum,14.5,19.17049,Gy(RBE),%
1,planned,Rectum,14.6,19.09786,Gy(RBE),%
1,planned,Rectum,14.7,19.04909,Gy(RBE),%
1,planned,Rectum,14.8,18.98888,Gy(RBE),%
1,planned,Rectum,34,9.50553,Gy(RBE),%
1,planned,Rectum,34.1,9.45993,Gy(RBE),%
1,planned,Rectum,34.2,9.42654,Gy(RBE),%
1,planned,Rectum,34.3,9.39345,Gy(RBE),%
1,planned,Rectum,34.4,9.35196,Gy(RBE),%
1,planned,Rectum,34.5,9.30604,Gy(RBE),%
1,planned,Rectum,34.6,9.27235,Gy(RBE),%
1,planned,Rectum,34.7,9.22334,Gy(RBE),%
1,planned,Rectum,34.8,9.18734,Gy(RBE),%
1,planned,Rectum,34.9,9.14867,Gy(RBE),%
1,planned,Rectum,35,9.11402,Gy(RBE),%
1,planned,Rectum,35.1,9.07618,Gy(RBE),%
1,planned,Rectum,35.2,9.04251,Gy(RBE),%
1,planned,Rectum,35.3,9.00141,Gy(RBE),%
1,planned,Rectum,35.4,8.96289,Gy(RBE),%
1,planned,Rectum,35.5,8.92638,Gy(RBE),%
1,planned,Rectum,35.6,8.89506,Gy(RBE),%
1,planned,Rectum,35.7,8.85644,Gy(RBE),%
1,planned,Rectum,35.8,8.81237,Gy(RBE),%
1,planned,Rectum,35.9,8.76545,Gy(RBE),%
1,planned,Rectum,36,8.73692,Gy(RBE),%
1,planned,Rectum,36.1,8.70149,Gy(RBE),%
1,planned,Rectum,36.2,8.66073,Gy(RBE),%
1,planned,Rectum,36.3,8.61303,Gy(RBE),%
1,planned,Rectum,36.4,8.56549,Gy(RBE),%
1,planned,Rectum,36.5,8.51527,Gy(RBE),%
1,planned,Rectum,36.6,8.47214,Gy(RBE),%
1,planned,Rectum,36.7,8.41663,Gy(RBE),%
1,planned,Rectum,36.8,8.37863,Gy(RBE),%
1,planned,Rectum,36.9,8.35041,Gy(RBE),%
1,planned,Rectum,37,8.31595,Gy(RBE),%
1,planned,Rectum,37.1,8.288,Gy(RBE),%
1,planned,Rectum,37.2,8.26272,Gy(RBE),%
1,planned,Rectum,37.3,8.23171,Gy(RBE),%
1,planned,Rectum,37.4,8.19804,Gy(RBE),%
1,planned,Rectum,37.5,8.1594,Gy(RBE),%
1,planned,Rectum,37.6,8.11729,Gy(RBE),%
1,planned,Rectum,37.7,8.06844,Gy(RBE),%
1,planned,Rectum,37.8,8.02818,Gy(RBE),%
1,planned,Rectum,37.9,7.96257,Gy(RBE),%
1,planned,Rectum,38,7.90243,Gy(RBE),%
1,planned,Rectum,38.1,7.84717,Gy(RBE),%
1,planned,Rectum,38.2,7.80889,Gy(RBE),%
1,planned,Rectum,38.3,7.77623,Gy(RBE),%
1,planned,Rectum,38.4,7.74385,Gy(RBE),%
1,planned,Rectum,38.5,7.71867,Gy(RBE),%
1,planned,Rectum,38.6,7.70076,Gy(RBE),%
1,planned,Rectum,38.7,7.6754,Gy(RBE),%
1,planned,Rectum,38.8,7.64753,Gy(RBE),%
1,planned,Rectum,38.9,7.59392,Gy(RBE),%
1,planned,Rectum,39,7.53856,Gy(RBE),%
1,planned,Rectum,39.1,7.4879,Gy(RBE),%
1,planned,Rectum,39.2,7.4423,Gy(RBE),%
1,planned,Rectum,39.3,7.40429,Gy(RBE),%
1,planned,Rectum,39.4,7.35858,Gy(RBE),%
1,planned,Rectum,39.5,7.30843,Gy(RBE),%
1,planned,Rectum,39.6,7.25325,Gy(RBE),%
1,planned,Rectum,39.7,7.22353,Gy(RBE),%
1,planned,Rectum,39.8,7.19164,Gy(RBE),%
1,planned,Rectum,39.9,7.16789,Gy(RBE),%
1,planned,Rectum,40,7.13184,Gy(RBE),%
1,planned,Rectum,40.1,7.09953,Gy(RBE),%
1,planned,Rectum,40.2,7.04322,Gy(RBE),%
1,planned,Rectum,40.3,6.98051,Gy(RBE),%
1,planned,Rectum,40.4,6.93635,Gy(RBE),%
1,planned,Rectum,40.5,6.90025,Gy(RBE),%
1,planned,Rectum,40.6,6.87001,Gy(RBE),%
1,planned,Rectum,40.7,6.83943,Gy(RBE),%
1,planned,Rectum,40.8,6.81393,Gy(RBE),%
1,planned,Rectum,40.9,6.7731,Gy(RBE),%
1,planned,Rectum,41,6.74696,Gy(RBE),%
1,planned,Rectum,41.1,6.71209,Gy(RBE),%
1,planned,Rectum,41.2,6.64682,Gy(RBE),%
1,planned,Rectum,41.3,6.5857,Gy(RBE),%
1,planned,Rectum,41.4,6.53214,Gy(RBE),%
1,planned,Rectum,41.5,6.48609,Gy(RBE),%
1,planned,Rectum,41.6,6.44336,Gy(RBE),%
1,planned,Rectum,41.7,6.3864,Gy(RBE),%
1,planned,Rectum,41.8,6.33488,Gy(RBE),%
1,planned,Rectum,41.9,6.30537,Gy(RBE),%
1,planned,Rectum,42,6.28613,Gy(RBE),%
1,planned,Rectum,42.1,6.27749,Gy(RBE),%
1,planned,Rectum,42.2,6.26234,Gy(RBE),%
1,planned,Rectum,42.3,6.23083,Gy(RBE),%
1,planned,Rectum,42.4,6.18859,Gy(RBE),%
1,planned,Rectum,42.5,6.12637,Gy(RBE),%
1,planned,Rectum,50,2.63461,Gy(RBE),%
1,planned,Rectum,50.1,2.61684,Gy(RBE),%
1,planned,Rectum,50.2,2.55227,Gy(RBE),%
1,planned,Rectum,50.3,2.48541,Gy(RBE),%
1,planned,Rectum,50.4,2.46586,Gy(RBE),%
1,planned,Rectum,50.5,2.39354,Gy(RBE),%
1,planned,Rectum,50.6,2.33448,Gy(RBE),%
1,planned,Rectum,50.7,2.28168,Gy(RBE),%
1,planned,Rectum,50.8,2.25787,Gy(RBE),%
1,planned,Rectum,50.9,2.19108,Gy(RBE),%
1,planned,Rectum,51,2.12473,Gy(RBE),%
1,planned,Rectum,51.1,2.11024,Gy(RBE),%
1,planned,Rectum,51.2,2.03551,Gy(RBE),%
1,planned,Rectum,51.3,1.98004,Gy(RBE),%
1,planned,Rectum,51.4,1.92951,Gy(RBE),%
1,planned,Rectum,51.5,1.89144,Gy(RBE),%
1,planned,Rectum,51.6,1.82465,Gy(RBE),%
1,planned,Rectum,51.7,1.77709,Gy(RBE),%
1,planned,Rectum,51.8,1.71624,Gy(RBE),%
1,planned,Rectum,51.9,1.65075,Gy(RBE),%
1,planned,Rectum,52,1.61509,Gy(RBE),%
1,planned,Rectum,52.1,1.58169,Gy(RBE),%
1,planned,Rectum,52.2,1.52462,Gy(RBE),%
1,planned,Rectum,52.3,1.44352,Gy(RBE),%
1,planned,Rectum,52.4,1.39243,Gy(RBE),%
1,planned,Rectum,52.5,1.34659,Gy(RBE),%
1,planned,Rectum,52.6,1.33099,Gy(RBE),%
1,planned,Rectum,52.7,1.27496,Gy(RBE),%
1,planned,Rectum,52.8,1.23031,Gy(RBE),%
1,planned,Rectum,52.9,1.15298,Gy(RBE),%
1,planned,Rectum,53,1.0894,Gy(RBE),%
1,planned,Rectum,53.1,1.05667,Gy(RBE),%
1,planned,Rectum,53.2,1.03679,Gy(RBE),%
1,planned,Rectum,53.3,1.00334,Gy(RBE),%
1,planned,Rectum,53.4,0.92593,Gy(RBE),%
1,planned,Rectum,53.5,0.85545,Gy(RBE),%
1,planned,Rectum,53.6,0.81901,Gy(RBE),%
1,planned,Rectum,53.7,0.77809,Gy(RBE),%
1,planned,Rectum,53.8,0.75188,Gy(RBE),%
1,planned,Rectum,57.6,NaN,Gy(RBE),%
1,planned,Rectum,57.7,NaN,Gy(RBE),%
1,planned,Rectum,57.8,NaN,Gy(RBE),%
1,planned,Rectum,57.9,NaN,Gy(RBE),%
1,planned,Rectum,58,NaN,Gy(RBE),%
1,planned,Rectum,58.1,NaN,Gy(RBE),%
1,planned,Rectum,58.2,NaN,Gy(RBE),%
1,planned,Rectum,58.3,NaN,Gy(RBE),%

【问题讨论】：

这可能是导致 KeyError 的原因：- (VolValues[loop] > 2) and (VolValues[loop+1] < 2)。在第一次迭代中，loop 的值是0，然后在下一行你访问VolValues[loop]，它变成VolValues[0]，这就是回溯中最后一行所说的，KeyError: 0 .因为VolValues 没有名为0 的密钥。
在我将它定义为一个函数之前它已经工作了......你似乎是正确的；我刚刚使用更少的代码复制了错误。 VolValues[23332] 访问第一个值，而不是 [0]。如何重置索引而不是让熊猫从更大的数组中保留它们？或者，我应该如何处理 pandas 中的数据子集？（在这种情况下，给定一个包含许多患者和来自两种方法的数据的大型“整理表”，仅访问来自给定方法的给定患者的数据。）
VolValues 是 list 还是 dictionary？
VolValues 是“pandas.core.series 模块的系列对象”。
我对熊猫没有任何经验。我已经对你的问题投了赞成票，希望它会得到更多关注。

标签： python pandas function numpy dataframe

【解决方案1】：

当 pandas 有一行没有索引 0 或数据框没有名为 0 的列时，会发生 Key Error:0。

在你的情况下，它是系列。考虑这个例子

df = pd.DataFrame({'vl':[1,2,3,4],'bh':[5,6,4,7]},index=[10,11,12,13])

df['bh'][0] #<-- leads to key error zero as the index doesn't contain 0.

因此，您可以将其更改为df['bh'].iloc[0]，它将返回5，或者您可以将其更改为df['bh'].values[0]，返回相同的值。

在您的情况下，它将是 VolValues.iloc[loop] 或 VolValues.values[loop]

【讨论】：

【解决方案2】：

VolValues[loop] 可能不想将零作为索引，很可能您需要从1（一）开始。大致如下：

for loop in range(1,len(VolValues)):

【讨论】：

正如我上面评论的那样， VolValues[23332] 访问第一个值（在患者 1 的计划数据的情况下），所以我认为你的建议不会奏效。显然，当我抓取一个大型 CSV 导入数据帧的相关切片时，它会保留相对于该数据帧的索引，而不是重新编号行。我想我尝试附加 .copy() 来尝试创建一个正确索引的新数组，但它没有更改错误消息，但我将在几个小时后再次检查。
您是否尝试将索引重置到位？
实际上是的，技术上不是：Redditor 'Nikota' 建议使用'values'，它通过将数据值放入一个新数组中忽略原始数组的索引来有效地做到这一点。另一个Redditor建议了一种在重置索引时保留索引列的方法，但我不需要原始表的索引，所以我没有使用这种方法。
@DBinJP 作为 JP 中的一个人与 JP 中的另一个人，我建议您加载交互式 python 提示并手动处理您的数据，尝试不同的索引并寻找正确的范围.我打赌你可以在几分钟内解决你的问题，绝对不到一个小时 =)

【解决方案3】：

Reddit user Nikota commented 仅提取值而不保留索引信息，就像我试图做的那样，只需使用.values 后缀。这似乎解决了我的问题。

因此，以下代码似乎可以按预期工作：

# Import Rectum DVH
import pandas
import numpy
df = pandas.read_csv('/Users/daniel/Documents/data/DVH/RectumData.csv',
                 delimiter=',',header=0)
# Calculate D_2%, defined by ICRU 78 as "the greatest dose which all but
# 2 percent of a [volume of interest] receives." aka D_{near-max}    
def interpD2(disttype):
# Loop through all patients' plans.
    Dose2Results = numpy.zeros(40)
    for num in range(0,40): 
# We know a priori that there is no DVH data with Volume = 2. Hence we look for
# the two columns less than and greater than Volume = 2.
        DoseValues = df.loc[(df['StudyID'] == num+1) & (df['DistributionType'] == disttype)].Dose.values
        VolValues = df.loc[(df['StudyID'] == num+1) & (df['DistributionType'] == disttype)].Volume.values
        for loop in range(0,len(VolValues)):
            if (VolValues[loop] > 2) and (VolValues[loop+1] < 2):
                LowerVolumeIndex,UpperVolumeIndex = loop,loop+1
                x0,x1,x2 = 2,VolValues[LowerVolumeIndex],VolValues[UpperVolumeIndex]
                y1,y2 = DoseValues[LowerVolumeIndex],DoseValues[UpperVolumeIndex]
                Dose2Results[num] = y1 - ((x1-x0)/(x2 - x1))*(y2 - y1)
    return Dose2Results
D2Planned = interpD2('planned')
D2Blurred = interpD2('blurred')

【讨论】：

或者您可以将其更改为VolValues.iloc[loop]，基本上您访问的是系列中不存在的行。如果您希望常规索引像数组一样工作，那么您必须选择.iloc。