如何在pandas中保存每个循环后的。csv中的输出而不重写？

Question

此外，我希望在打印输出时跳过零值行。

这是我的代码：

import pandas as pd `#pandas library
sample = pd.DataFrame(pd.read_csv ("Sample.csv")) #importing .csv as pandas DataFrame

i = 0
while (i <= 23):
    print('Value for', i) `#i vale`
    sample2 = (sample[sample['Hour'] == i])`#Data for every hour`
    sample3 = (sample2[(sample2['GHI']) == (sample2['GHI'].max(0))]) `#Max value from sample3 DataFrame`
    sample3 = sample3.loc[sample3.ne(0).all(axis=1)]`ignoring all rows having zero values`
    print(sample3)  `print sample3`
    sample3.to_csv('Output.csv')`trying to save for output after every iteration`
    i = i + 1

Answer 1

做你想做的事情的另一种方法是摆脱你的循环，如下所示：

sample_with_max_ghi = sample.assign(max_ghi=sample.groupby('Hour')['GHI'].transform('max'))
sample_filtered = sample_with_max_ghi[sample_with_max_ghi['GHI'] == sample_with_max_ghi['max_ghi']]
output_sample = sample_filtered.loc[sample_filtered.ne(0).all(axis=1)].drop('max_ghi', axis=1)
output_sample.to_csv('Output.csv')

一些解释：

1.

sample_with_max_ghi = sample.assign(max_ghi=sample.groupby('Hour')['GHI'].transform('max'))

sample_filtered = sample_with_max_ghi[sample_with_max_ghi['GHI'] == sample_with_max_ghi['max_ghi']]

output_sample = sample_filtered.loc[sample_filtered.ne(0).all(axis=1)].drop('max_ghi', axis=1)