How to open org-mode table as a pandas dataframe (with column names)?

Question

I can't find how to open an org-mode table as a pandas dataframe.

Here is my example:

#+NAME:tbl
| a | b |
|---|---|
| 1 | 2 |

#+BEGIN_SRC python :var tbl=electro :results replace output   :session test :exports both
import pandas as pd
df = pd.DataFrame(tbl)
print(df)
#+END_SRC

Here I get the dataframe, but the columns are 1 2

Answer 1

I am sorry, I found (probably) a solution (my actual example):

#+NAME:electro
| TIME                    |   low | high | heatpump | water | accurain | repy1 | repy2 | repy3 | repy4 |
|-------------------------+-------+------+----------+-------+----------+-------+-------+-------+-------|
| 2016-10-15 08:41:00.000 | 24830 | 3242 |     7308 |  2294 |      NaN |       |       |       |       |
| 2016-10-23 10:38:53.000 | 24963 | 3254 |     7362 |  2317 |          |       |       |       |       |
| 2016-10-30 19:36:38.000 | 25107 | 3265 |     7417 |  2342 |          |       |       |       |       |
|


#+BEGIN_SRC python :var tbl=electro :hlines no  :colnames no :results replace output   :session test :exports both
import pandas as pd
df = pd.DataFrame(tbl[1:], columns=tbl[0])

# and now more tricks...

df = df.replace('', np.nan)
df = df.replace('NaN', np.nan)
print(df)
df.to_hdf("test.h5",key="key1",mode='w')
#+END_SRC

The solution was to add :colnames no to prevent babel from removing the header.