how to access timestamp in DataFrame pandas?

Question

Having downloaded data from yahoo for a stock using the get_data_yahoo I then want to access the time for each row... How do I do that?

One way I've kind of figured out to do this is:

it = stock.iterrows()
st0 = it.next()
resultIWant = st0[0].value # this gives what I want (almost)
print resultIWant

EDIT:

So basically I want something like stock['Open'] but for time, smth like stock['Time'] would be ideal..

some one asked for the output of .head

            Open  High   Low  Close  Volume  Adj Close
Date                                                  
2012-04-03  2.06  2.09  2.01   2.08  463200       2.08
2012-04-04  2.04  2.05  2.01   2.02  335600       2.02

Expected output from function:

print find_time(stock,2) # function I'm looking for
1333497600000000000      # resulting output

The expected output is the time from the last epoch for each of the dates in an array or some other way to get the time of each entry. The example code I gave gives me the results I want however if I want to get access to the 4th element the only way to do it would be to .next the iterator 4 times, this seems like a bad method.

First Question: Is there a more obvious way, what I'm doing doesn't seem like the best way to do this.

Second Question: What units is the result in? I think it's nanoseconds but not sure...

What do you want to do with these times, ultimately? Help us understand where you're going with this, and you'll get a clearer answer. — Dan Allan
– Dan Allan, Commented Apr 3, 2013 at 15:54
I want to subtract subtract stockTime[1] - stockTime[0] so I need the value of the first and the second element... ultimately i'm trying to build a candlestickchart and want to have the right widths for the bars — evan54
– evan54, Commented Apr 3, 2013 at 18:00

Andy Hayden · Accepted Answer · 2013-04-03 16:12:16Z

3

Timestamps have a time method:

In [1]: t = pd.Timestamp('200101011300')

In [2]: t
Out[2]: <Timestamp: 2001-01-01 13:00:00>

In [3]: t.time()
Out[3]: datetime.time(13, 0)

The value is nanoseconds since midnight 1 January 1970 i.e. unix time * 10 ** 9:

In [4]: t.value
Out[4]: 978354000000000000

edited Apr 3, 2013 at 16:12

answered Apr 3, 2013 at 16:05

Andy Hayden

378k110 gold badges640 silver badges546 bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

evan54 Over a year ago

hm.... but my question is with regards to how I get the Timestamp part of the stock out? I have no idea how to access it other than what I showed..

Andy Hayden Over a year ago

@evan54 it's not entirely clear on what the structure is of your DataFrame, an example would help a lot, e.g. the output of df.head(). Certainly you want to avoid using a for loop.

evan54 Over a year ago

the output I expect is the same as from the code I've posted. I updated the question to include .head output and the expected output of the function I'm after... Hopefully that clears things up

Andy Hayden Over a year ago

Still not sure what you're asking, but if you have a function for one value, you can then use Series/DataFrame apply or index's map, e.g. df.index.map(lambda t: t.value) ?

MasterKZ · Accepted Answer · 2020-06-19 14:46:06Z

3

Just seen the question. But may be it is interesting to somebody.

In this case you can make next command:

stock['Time'] = stock.index

then you'll get new column 'Time' cause your Date is the index. To convert the DateTime to Unix Time change to:

import numpy as np

stock['Time'] = stock.index.astype(np.int64).

edited Jun 19, 2020 at 14:46

answered Jun 19, 2020 at 13:37

MasterKZ

312 bronze badges

Comments

evan54 · Accepted Answer · 2013-04-13 16:51:38Z

2

The correct answer I found in the end here

You can access the rows using st.ix[index_val] from section 5.2.8 of the reference.

ts = st.ix[2].Name

which returns a timestamp

ts is the same as the st[0] in the question

answered Apr 13, 2013 at 16:51

evan54

3,7537 gold badges39 silver badges66 bronze badges

Comments

waitingkuo · Accepted Answer · 2013-04-03 16:37:29Z

0

In fact each row contains a tuple contains index and data, so you might do this:

for index, row in stock.iterrows():
    print index

answered Apr 3, 2013 at 16:37

waitingkuo

94.5k28 gold badges119 silver badges122 bronze badges

3 Comments

evan54 Over a year ago

any way of accessing it out of a loop? I mean this is pretty much what I showed right? I want to get stock['Time'] or something like that..

waitingkuo Over a year ago

What output do you expect?

evan54 Over a year ago

I posted what I expect, it is basically what the .next implementation outputs

Collectives™ on Stack Overflow

how to access timestamp in DataFrame pandas?

4 Answers 4

4 Comments

Comments

Comments

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

4 Comments

Comments

Comments

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related