我尝试做的是在csv中读取，根据特定列消除重复值，然后将数据减少到不再比15的增量更接近点。
我的代码在文件中读取正常，然后drop_duplicates按需要工作，然后我按该列对其进行排序。为了减少数据，我创建了一个新数据框，其中包含现有数据的第一行，然后我将浏览相关列中的每个值，并将其附加到新数据框中＆＃ 39;比对照值高至少15千克/小时。
我的数据框架没有正确组合，我最后得到的结果数据框如下所示：
Unnamed: 0 0
TimeStamp (s) 0.002
TC 01 (C) 30.6689
TC 02 (C) 28.6879
TC 03 (C) 27.9779
TC 32 (C) 22.6416
Product Back Pressure (kPa) 0.166353
Product Mass Flow (kg/hr) 107.427
Semtech Flow (kg/hr) 28.2135
Mass Flow (kg/hr) 28.2135
Voltage (V) 1.63065
Angle (degrees) 0
1 Unnamed: 0 TimeStamp (s) TC 01 (C) TC 02...
2 Unnamed: 0 TimeStamp (s) TC 01 (C) TC 0...
3 Unnamed: 0 TimeStamp (s) TC 01 (C) TC 02...
4 Unnamed: 0 TimeStamp (s) TC 01 (C) TC 02...
我显然做错了什么，但至少比我试图在＆＃34;＆＃34;＆＃34;声明。
def import_df():
new_df = pd.read_csv(os.path.join(pathname, f), delimiter = ',')
new_df = new_df.drop_duplicates(subset = 'Mass Flow (kg/hr)')
new_df = new_df.sort_values('Mass Flow (kg/hr)')
reduced_df = new_df.iloc[0]
current_mass_flow = new_df['Mass Flow (kg/hr)'].iloc[0]
i = 1
for value in new_df['Mass Flow (kg/hr)']:
if value < current_mass_flow + 15:
reduced_df.loc[i] = new_df.loc[new_df['Mass Flow (kg/hr)'] == value]
current_mass_flow = value
i += 1
else: next
return reduced_df
我该怎么做才能纠正这个问题？它显然没有按照我期望的方式添加到数据框中。我肯定错过了一些关于如何向这个数据帧添加行的更好的观点。
此外，我无法帮助，但我觉得有一种更简单/更直接的方式来完成我想要做的事情。
Sample Source Data
TimeStamp (s) TC 01 (C) TC 02 (C) TC 03 (C) TC 32 (C) Product Back Pressure (kPa) Product Mass Flow (kg/hr) Semtech Flow (kg/hr) Mass Flow (kg/hr) Voltage (V) Angle (degrees)
0 0.004 493.2881108 296.1245877 255.8202916 26.3430426 0.297276487 147.4692621 30.21243527 30.21243527 1.634457337 0
1 0.178 493.2881108 296.1245877 255.8202916 26.3430426 0.283227103 147.4692621 30.21243527 30.21243527 1.634457337 0
2 1.178 493.1325481 296.155699 255.8514043 26.3430426 0.283227103 144.5363918 31.06903075 31.06903075 1.634457337 0
3 2.178 493.0703231 296.2490329 255.8825171 26.3430426 0.289335716 141.244467 31.06903075 31.06903075 1.634457337 0
4 3.178 492.4480726 296.373478 255.8825171 26.40525146 0.292389668 141.244467 29.73651711 29.73651711 1.634139868 0
5 4.178 493.2881108 296.373478 255.9136299 26.3430426 0.292389668 146.0926428 30.40291693 30.40291693 1.634457337 0
6 5.178 493.2881108 296.4357006 255.8825171 26.40525146 0.289742626 146.0926428 30.40291693 30.40291693 1.634457337 0
7 6.178 492.8836479 296.4045893 255.9136299 26.40525146 0.281191135 146.0926428 30.78359426 30.78359426 1.634139868 0
8 7.178 493.1325481 296.373478 255.9447427 26.40525146 0.281191135 146.2123624 30.02223961 30.02223961 1.634457337 0
90 959.629 442.3250036 300.5424521 264.6564452 27.77387677 0.593127726 203.9719224 44.39531112 44.39531112 1.635409746 0
91 960.629 442.231666 300.5424521 264.6564452 27.77387677 0.599643603 203.9719224 44.77598845 44.77598845 1.634457337 0
92 961.629 441.3605153 300.3557844 264.6564452 27.77387677 0.58966651 199.4828012 44.77598845 44.77598845 1.634457337 0
93 962.629 441.0493901 300.3557844 264.6253324 27.77387677 0.58966651 199.1237467 43.63367047 43.63367047 1.634774807 0
94 963.629 441.0493901 300.1691166 264.531994 27.77387677 0.58885198 199.1237467 43.63367047 43.63367047 1.635092276 0
95 964.629 441.2360652 300.3868956 264.531994 27.77387677 0.588444716 203.8522028 43.63367047 43.63367047 1.635092276 0
96 965.629 441.4849654 300.3557844 264.4697685 27.77387677 0.588444716 199.1237467 43.63367047 43.63367047 1.634139868 0
97 966.629 441.3916279 300.2935618 264.4697685 27.77387677 0.597403826 199.1237467 44.39531112 44.39531112 1.633823352 0
98 967.629 441.7338656 300.4802295 264.531994 27.77387677 0.592720461 203.8522028 44.39531112 44.39531112 1.634139868 0
99 968.629 441.2982903 300.6046747 264.6253324 27.77387677 0.592720461 203.9719224 43.63367047 43.63367047 1.634139868 0
100 969.629 441.578303 300.6980086 264.687558 27.77387677 0.606769845 203.9719224 45.06142494 45.06142494 1.634139868 0
101 970.629 441.8894282 300.5735634 264.687558 27.77387677 0.594145709 200.3806463 45.06142494 45.06142494 1.635092276 0
期望输出
TimeStamp (s) TC 01 (C) TC 02 (C) TC 03 (C) TC 32 (C) Product Back Pressure (kPa) Product Mass Flow (kg/hr) Semtech Flow (kg/hr) Mass Flow (kg/hr) Voltage (V) Angle (degrees)
13 12.178 493.008098 296.2490329 255.8825171 26.3430426 0.31682341 146.0327308 29.26059896 29.26059896 1.634139868 0
77 947.156 443.7872922 301.3202954 264.9986859 27.74277234 0.613081913 199.8419601 44.39531112 44.39531112 1.637947595 0.158889819
答案 0 :(得分：2)
检查此代码，如果您正在寻找此信息，请与我们联系。 pandas.DataFrame.iterrows用于循环记录并检查质量流量。
pandas.DataFrame.iterrows ---一个迭代帧的行的生成器。
import pandas as pd
import os
new_df = pd.read_csv(os.path.join('C:\Shijo\Python\sample.txt'), delimiter = ',')
new_df = new_df.drop_duplicates(subset = 'Mass Flow (kg/hr)')
new_df = new_df.sort_values('Mass Flow (kg/hr)')
new_df = new_df.reset_index(drop=True)
current_mass_flow = new_df.iloc[0]['Mass Flow (kg/hr)']
indexlst=[1]
for index, row in new_df.iterrows():
if row['Mass Flow (kg/hr)'] > current_mass_flow + 15:
print ("Mathcing index : ",index)
indexlst.append(index)
current_mass_flow =row['Mass Flow (kg/hr)']
reduced_df= new_df.iloc[indexlst]
print (reduced_df )
输出
TimeStamp (s) TC 01 (C) TC 02 (C) TC 03 (C) TC 32 (C) Product Back Pressure (kPa) Product Mass Flow (kg/hr) Semtech Flow (kg/hr) Mass Flow (kg/hr) Voltage (V) Angle (degrees)
1 7.178 493.132548 296.373478 255.944743 26.405251 0.281191 146.212362 30.022240 30.022240 1.634457 0
8 960.629 442.231666 300.542452 264.656445 27.773877 0.599644 203.971922 44.775988 44.775988 1.634457 0
答案 1 :(得分：1)
您好像在寻找pandas.cut，您可以在其中指定垃圾箱（例如np.arange）。例如，请参阅this question。
编辑：
使用groupby和cut
new_df.groupby(pd.cut(new_df['Mass Flow (kg/hr)'],\
np.arange(new_df['Mass Flow (kg/hr)'].min(),
new_df['Mass Flow (kg/hr)'].max()+1,
15)))\
.apply(lambda x: x.loc[x['Mass Flow (kg/hr)'].idxmin()])
答案 2 :(得分：0)
来自重复删除的公寓，将您的数据拆分为“可重现”的垃圾箱会不会很有趣？
df['bins'] = df['Mass Flow (kg/hr)'] // 15 # Create bins from 0
df.groupby(['bins'], as_index=False).mean().drop('bins', axis=1) # Get mean values
# and remove "bin" column
否则，您可以针对系列的最小值应用与规范化相同的过程。
df['bins'] = (df['Mass Flow (kg/hr)'] - df['Mass Flow (kg/hr)'].min()) // 15 # Normalized bins
df.groupby(['bins'], as_index=False).mean().drop('bins', axis=1)
请注意，此代码已直接在您的数据上进行测试（无法使用pandas.read_clipboard()），但适用于“类似”数据框。