python performance problems using loops with big tables

问题

I am using python and multiple libaries like pandas and scipy to prepare data so I can start deeper analysis. For the preparation purpose I am for instance creating new columns with the difference of two dates.
My code is providing the expected results but is really slow so I cannot use it for a table with like 80K rows. The run time would take ca. 80 minutes for the table just for this simple operation.

The problem is definitely related with my writing operation:

tableContent[6]['p_test_Duration'].iloc[x] = difference

Moreover python is providing a Warning:

complete code example for date difference:

import time
from datetime import date, datetime

tableContent[6]['p_test_Duration'] = 0

#for x in range (0,len(tableContent[6]['p_test_Duration'])):
for x in range (0,1000):
    p_test_ZEIT_ANFANG = datetime.strptime(tableContent[6]['p_test_ZEIT_ANFANG'].iloc[x], '%Y-%m-%d %H:%M:%S')
    p_test_ZEIT_ENDE = datetime.strptime(tableContent[6]['p_test_ZEIT_ENDE'].iloc[x], '%Y-%m-%d %H:%M:%S')
    difference = p_test_ZEIT_ENDE - p_test_ZEIT_ANFANG

    tableContent[6]['p_test_Duration'].iloc[x] = difference

the correct result table:

回答1:

Take away the loop, and apply the functions to the whole series.

ZEIT_ANFANG = tableContent[6]['p_test_ZEIT_ANFANG'].apply(lambda x: datetime.strptime(x, '%Y-%m-%d %H:%M:%S'))
ZEIT_ENDE = tableContent[6]['p_test_ZEIT_ENDE'].apply(lambda x: datetime.strptime(x, '%Y-%m-%d %H:%M:%S'))
tableContent[6]['p_test_Duration'] = ZEIT_ENDE - ZEIT_ANFANG

回答2:

You can vectorize the conversion of dates by using pd.to_datetime and avoid using apply unnecessarily.

tableContent[6]['p_test_Duration'] = (
    pd.to_datetime(tableContent[6]['p_test_ZEIT_ENDE']) -
    pd.to_datetime(tableContent[6]['p_test_ZEIT_ANFANG'])
)

---

Also, you were getting the SettingWithCopy warning because of the chained indexing assingnment

tableContent[6]['p_test_Duration'].iloc[x] = difference

Which you don't have to worry about if you go about it in the way I suggested.

回答3:

The other answers are fine, but I would recommend that you avoid chained indexing in general. The pandas docs explicitly discourage chained indexing as it either produces unreliable results or is slow (due to multiple calls to __getitem__). Assuming your data frame is multi-indexed, you might replace:

tableContent[6]['p_test_Duration'].iloc[x] = difference

with:

tableContent.loc[x, (6, 'p_test_Duration')] = difference

You can sometimes get around this issue, but why not learn the method least likely to cause problems in the future?

来源：https://stackoverflow.com/questions/45057978/python-performance-problems-using-loops-with-big-tables