How to convert that html table to pandas dataframe?
https://www.google.com/finance/getprices?q=HINDALCO&i=60&p=15d&f=d,o,h,l,c,v
Example data:
Asked
Active
Viewed 1,118 times
1
-
1BTW, pandas comes with a data reader for [google finance](http://pandas-datareader.readthedocs.io/en/latest/remote_data.html#remote-data-google). Maybe it does the job? – greole Apr 11 '17 at 06:25
3 Answers
2
You can use read_csv with parameters skiprows and names for new column names:
url = 'https://www.google.com/finance/getprices?q=HINDALCO&i=60&p=15d&f=d,o,h,l,c,v'
df = pd.read_csv(url, skiprows=[0,1,2,3,5,6]).rename(columns={'COLUMNS=DATE':'DATE'})
print (df.head())
DATE CLOSE HIGH LOW OPEN VOLUME
0 a1490154300 194.80 194.80 194.80 194.80 2600
1 1 193.55 194.70 193.00 194.15 339142
2 2 193.80 193.95 193.55 193.60 242687
3 3 194.20 194.40 193.80 193.90 119874
4 4 193.80 194.20 193.80 194.20 121355
jezrael
- 822,522
- 95
- 1,334
- 1,252
-
do you any idea about datetime how can i convert that date[1]mean 1min date 2 mean 2 mins so on – Lakhan Pahuja Apr 11 '17 at 08:10
-
-
can i make array of time ? date 'a1490154300' mean 9:15 date'1'mean 9:16 so on – Lakhan Pahuja Apr 11 '17 at 15:26
-
Unfortunately I am offline now, only on phone, so fastest is create new question. Dont forget add sample data, desired output and what you try. Btw, in `a1490154300` is uncoded also date like `2017-04-11`? `a149015` means 9:15? The best is ecplain it too. Thanks. – jezrael Apr 11 '17 at 16:00
0
If you are not particular about getting your data only from Google, this will work with Yahoo :
import pandas as pd
import pandas_datareader.data as web
from datetime import datetime
start = datetime(2014, 6, 2)
end = datetime(2014, 9, 5)
hindalco = web.DataReader('HINDALCO.NS', 'yahoo', start, end)
In [15]: hindalco.head(5)
Out[15]:
Open High Low Close Volume Adj Close
Date
2014-06-02 147.3 151.15 146.35 150.05 13844600 146.2012
2014-06-03 150.7 155.70 149.35 155.20 23276100 151.2191
2014-06-04 156.0 161.75 155.20 160.70 15948200 156.5780
2014-06-05 160.6 171.00 159.95 169.85 20296900 165.4934
2014-06-06 172.0 172.10 165.30 169.25 13769100 164.9087
You will need to install pandas-datareader.
sudo -H pip install pandas-datareader (ubuntu)
pip install pandas-datareader (windows)
dmdip
- 1,665
- 14
- 15
-
I did try using 'HINDALCO' as the company code and google instead of yahoo, but this did not work. However, US stocks like Apple worked with Google : apple = web.DataReader('AAPL', 'google', start, end) – dmdip Apr 11 '17 at 06:36
-
0
@jezrael] has answered your question for one minute data for Hindalco. Here's more or less the same answer, tweaked a bit :
import pandas as pd
period = 60 #one minute
days = 2 # 2 days of data
ticker = 'HINDALCO'
url = 'http://www.google.com/finance/getprices?i={}&p={}d&f=d,o,h,l,c,v&df=cpct&q={}'.format(period, days, ticker)
cols = ['minute', 'open', 'high', 'low', 'close', 'volume']
df = pd.read_csv(url, skiprows=8, header=None, names=cols)
df.head()
Output:
minute open high low close volume
0 1 194.35 194.35 193.80 193.85 25785
1 2 194.15 194.40 194.00 194.35 64580
2 3 193.95 194.25 193.85 194.15 42980
3 4 193.80 193.95 193.75 193.90 33936
4 5 193.90 193.90 193.60 193.80 57088
dmdip
- 1,665
- 14
- 15
-
Yes, you can. You have to use the field value which appears as **a1490154300**. Basically 1490154300 is the start time of the data, measured as the number of seconds from January 1, 1970 5.30 AM IST. (Actually 0000 UTC). To that field value add the number of minutes which appear in that column. – dmdip Apr 11 '17 at 17:21
-
Correction to the code given earlier. df = pd.read_csv(url, skiprows=7, header=None, names=cols) – dmdip Apr 11 '17 at 17:33