Filtering Python Dataframe by Minutes

Hi Developer Community!

I have filtered my minute time series data frame by specific hours and minutes. The first Loop filters for 11:30 the second for 9:30. Lastly, I concatenated them next to each other.

Is there a way to do the same procedure without using an extra (second) loop? I would like to add more time and hour filters.

Thank you

Find more posts tagged with

python

filter

time-series

pandas

Accepted answers

Jirapongse

@levente.letenyei

Yes, it is feasible to do it in one loop.

The code looks like this:

for i in range(len(majors_list)):
    temp = df_fixing[i][(df_fixing[i].index.hour==11) & (df_fixing[i].index.minute==30)]
    temp.index = temp.index.date
    df_fixing[i] = df_fixing[i][(df_fixing[i].index.hour==9) & (df_fixing[i].index.minute==30)]
    df_fixing[i].index = df_fixing[i].index.date    
    df_fixing[i] = pd.concat([df_fixing[i], temp], axis=1)

All comments

aramyan.h

Hi @levente.letenyei ,

One quick solution could simply be adding an OR component (df_fixing[i].index.hour == 9) in your first loop so the final code will look like this:

for i in range(len(majors_list)):
    empty_list.append(df_fixing[i][((df_fixing[i].index.hour==11) | (df_fixing[i].index.hour == 9)) & (df_fixing[i].index.minute == 30)])

If you have noticed I have got rid of empty_list[i].index = empty_list[i].index.date to have the full datetime in the final list. And here is the final output:

empty_list[0]

Hope this helps.

Best regards,

Haykaz

levente.letenyei

Thank you ver much for your answer @aramyan.h ,

however I need the filtered minutes/hours data next to each other in new columns just like on the screenshot. Do you think it is feasible with one loop?

Jirapongse

@levente.letenyei

Yes, it is feasible to do it in one loop.

The code looks like this:

for i in range(len(majors_list)):
    temp = df_fixing[i][(df_fixing[i].index.hour==11) & (df_fixing[i].index.minute==30)]
    temp.index = temp.index.date
    df_fixing[i] = df_fixing[i][(df_fixing[i].index.hour==9) & (df_fixing[i].index.minute==30)]
    df_fixing[i].index = df_fixing[i].index.date    
    df_fixing[i] = pd.concat([df_fixing[i], temp], axis=1)

levente.letenyei

Hi @Jirapongse ,

thank you very much for your answer! This way of solution worked completely! Cheers!

EXPLORE OUR SITES