Question

我有一个ascii文件，其中包含3行数据，如下所示：

Timestamp: 00:47:14 SATID 13 VAL1 28 VAL2 227 SIGNAL 37                                     SATID 15 VAL1 22 VAL2 265 SIGNAL 30 SATID 16 VAL1 22 VAL2 265 SIGNAL 30
Timestamp: 00:48:14 SATID 13 VAL1 28 VAL2 227 SIGNAL 37                                     SATID 15 VAL1 22 VAL2 265 SIGNAL nan SATID 16 VAL1 22 VAL2 265 SIGNAL 30
Timestamp: 00:49:14                                     SATID 14 VAL1 22 VAL2 265 SIGNAL 30

（请参阅图像了解原始格式）。 original ascii data format 当我尝试将其读入Python时，我收到以下错误：

time,sat1,sat2,sat3,sat4 = np.loadtxt("test1.asc", usecols=(1,9,17,25,33), unpack=True, converters = {1: strpdate2num("%H:%M:%S")})
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python/numpy/lib/npyio.py", line 839, in loadtxt
vals = [vals[i] for i in usecols]
IndexError: list index out of range

有谁知道我怎么做才能让Python忽略空单元格并读取每列中可用的数据？

谢谢！

Answer 1

没有抓住numpy或pandas，让我们看看我们将如何阅读这个＆＃34; 手动＆＃34;

首先要认识到时间戳始终位于同一位置，然后使用" SATID "，因此您可以.split(' SATID ')[0]获取该信息。

然后看来，如果您对信息的其余部分执行.split(' SATID ')，则会获得所有必需的信息，然后您可以进一步拆分。

代码中的

看起来像这样：

raw_data = ["Timestamp: 00:47:14 SATID 13 VAL1 28 VAL2 227 SIGNAL 37                                     SATID 15 VAL1 22 VAL2 265 SIGNAL 30 SATID 16 VAL1 22 VAL2 265 SIGNAL 30",
            "Timestamp: 00:48:14 SATID 13 VAL1 28 VAL2 227 SIGNAL 37                                     SATID 15 VAL1 22 VAL2 265 SIGNAL nan SATID 16 VAL1 22 VAL2 265 SIGNAL 30",
            "Timestamp: 00:49:14                                     SATID 14 VAL1 22 VAL2 265 SIGNAL 30"]

output = []
for line in raw_data:
    if 'SATID' in line:  #making sure it is not an empty line
        timestamp = line.split(' SATID ')[0].split('Timestamp: ')[1].rstrip(' ')
        data = line.split(' SATID ')[1:]
        for record in data:
            if 'VAL1' in record:  #making sure it is not an empty record
                satid = record.split(' VAL1 ')[0]
                val1 = record.split(' VAL1 ')[1].split(' VAL2 ')[0]
                val2 = record.split(' VAL2 ')[1].split(' SIGNAL ')[0]
                signal = record.split(' SIGNAL ')[1].rstrip(' ')
                output.append({'Timestamp':timestamp,
                               'SATID':satid,
                               'VAL1':val1,
                               'VAL2':val2,
                               'SIGNAL':signal})


# output is now a list of dictionaries
for d in output:
    print(d)

Answer 2

由于列边不相交，您可以将文件视为固定宽度文件并使用函数read_fwf。您必须准备列规范列表 - 指定每列的第一个和最后一个位置的元组列表。这是规格的开头（很无聊，但你只需要做一次）：

specs = [(0,11),(11,20),(20,26),(26,29),(29,33),(33,37),
         (37,42),(42,45),(45,52),(52,55),(55,61),(61,63)]
pd.read_fwf('foo.txt',header=None,colspecs=specs)
#            0         1      2     3     4     5     6      7       8     9   \
#0  Timestamp:  00:47:14  SATID  13.0  VAL1  28.0  VAL2  227.0  SIGNAL  37.0   
#1  Timestamp:  00:48:14  SATID  13.0  VAL1  28.0  VAL2  227.0  SIGNAL  37.0   
#2  Timestamp:  00:49:14    NaN   NaN   NaN   NaN   NaN    NaN     NaN   NaN   

#      10   11  
#0    NaN  NaN  
#1    NaN  NaN  
#2  SATID  1.0

读取丢失单元格的ascii数据行？

2 个答案: