我有一組 csv 文件，其中日期和時間作為前兩列（文件中沒有標題）。這些文件在 Excel 中打開得很好，但是當我嘗試使用 Pandas read_csv 將它們讀入 Python 時，無論我是否嘗試類型轉換，都只返回第一個日期。當我在記事本中打開時，它不僅僅是逗號分隔，而且在第 1 行之后的每一行之前都有大量空格；我已嘗試skipinitialspace = True無濟于事我也嘗試過各種類型轉換，但都不起作用。我目前正在使用parse_dates = [['Date','Time']], infer_datetime_format = True, dayfirst = True輸出示例（無轉換）： 0 1 2 3 4 ... 12 13 14 15 160 02/03/20 15:13:39 5.5 5.8 42.84 ... 30.0 79.0 0.0 0.0 0.01 NaN 15:13:49 5.5 5.8 42.84 ... 30.0 79.0 0.0 0.0 0.02 NaN 15:13:59 5.5 5.7 34.26 ... 30.0 79.0 0.0 0.0 0.03 NaN 15:14:09 5.5 5.7 34.26 ... 30.0 79.0 0.0 0.0 0.04 NaN 15:14:19 5.5 5.4 17.10 ... 30.0 79.0 0.0 0.0 0.0... ... ... ... ... ... ... ... ... ... ... ...39451 NaN 01:14:27 5.5 8.4 60.00 ... 30.0 68.0 0.0 0.0 0.039452 NaN 01:14:37 5.5 8.4 60.00 ... 30.0 68.0 0.0 0.0 0.039453 NaN 01:14:47 5.5 8.4 60.00 ... 30.0 68.0 0.0 0.0 0.039454 NaN 01:14:57 5.5 8.4 60.00 ... 30.0 68.0 0.0 0.0 0.039455 NaN NaN NaN NaN NaN ... NaN NaN NaN NaN NaN以及 parse_dates 等： Date_Time pH1 SP pH Ph1 PV pH ... 1 2 30 02/03/20 15:13:39 5.5 5.8 ... 0.0 0.0 0.01 nan 15:13:49 5.5 5.8 ... 0.0 0.0 0.02 nan 15:13:59 5.5 5.7 ... 0.0 0.0 0.03 nan 15:14:09 5.5 5.7 ... 0.0 0.0 0.04 nan 15:14:19 5.5 5.4 ... 0.0 0.0 0.0從記事本復制的數據（實際上每行前面有更多空格，但在這里不起作用）：數據來自67.csv
1 回答
慕勒3428872
TA貢獻1848條經驗 獲得超6個贊
該文件充滿了NUL, '\x00'，需要將其刪除。
清理行后，用于pandas.DataFrame從 加載數據。d
import pandas as pd
import string # to make column names
# the issue is the the file is filled with NUL not whitespace
def import_file(filename):
# open the file and clean it
with open(filename) as f:
d = list(f.readlines())
# replace NUL, strip whitespace from the end of the strings, split each string into a list
d = [v.replace('\x00', '').strip().split(',') for v in d]
# remove some empty rows
d = [v for v in d if len(v) > 2]
# load the file with pandas
df = pd.DataFrame(d)
# convert column 0 and 1 to a datetime
df['datetime'] = pd.to_datetime(df[0] + ' ' + df[1])
# drop column 0 and 1
df.drop(columns=[0, 1], inplace=True)
# set datetime as the index
df.set_index('datetime', inplace=True)
# convert data in columns to floats
df = df.astype('float')
# give character column names
df.columns = list(string.ascii_uppercase)[:len(df.columns)]
# reset the index
df.reset_index(inplace=True)
return df.copy()
# call the function
dfs = list()
filenames = ['67.csv']
for filename in filenames:
dfs.append(import_file(filename))
display(df)
A B C D E F G H I J K L M N O
datetime
2020-02-03 15:13:39 5.5 5.8 42.84 7.2 6.8 10.63 60.0 0.0 300.0 1.0 30.0 79.0 0.0 0.0 0.0
2020-02-03 15:13:49 5.5 5.8 42.84 7.2 6.8 10.63 60.0 0.0 300.0 1.0 30.0 79.0 0.0 0.0 0.0
2020-02-03 15:13:59 5.5 5.7 34.26 7.2 6.8 10.63 60.0 22.3 300.0 1.0 30.0 79.0 0.0 0.0 0.0
2020-02-03 15:14:09 5.5 5.7 34.26 7.2 6.8 10.63 60.0 15.3 300.0 45.0 30.0 79.0 0.0 0.0 0.0
2020-02-03 15:14:19 5.5 5.4 17.10 7.2 6.8 10.63 60.0 50.2 300.0 86.0 30.0 79.0 0.0 0.0 0.0
添加回答
舉報
0/150
提交
取消