Python dict with values as tuples to pandas DataFrame

Question

I have the following dict:

td = {'q1':(111,222), 'q2':(333,444)} 

I would like to convert it to a dataframe that looks like this:

Query    Value1     Value2 q1       111       222 q2       333       444 

I have tried the following:

df = pd.DataFrame(td.items()) 

The result looks like this:

    0         1 0   q1  (111,222) 1   q2  (333,444)  

If it wasn't entirely obvious, I am new to python and pandas. How can I get a dictionary with values as tuples to behave as separate columns in a dataframe?

My end goal is to display percent difference between value1 and value2.

Answer 1

Setup

td = {'q1':(111,222), 'q2':(333,444)} 

Option 1

pd.DataFrame(td).T.rename_axis('Query').add_prefix('Value').reset_index()    Query  Value0  Value1 0    q1     111     222 1    q2     333     444 

---

Option 2

from cytoolz.dicttoolz import merge  pd.DataFrame(     [merge(         {'Query': k},         {'Value{}'.format(i): x for i, x in enumerate(v, 1)}      ) for k, v in td.items()] )    Query  Value1  Value2 0    q1     111     222 1    q2     333     444 

---

Response to Comment

df = pd.DataFrame(td).T.rename_axis('Query').add_prefix('Value') df.assign(PctChg=df.pct_change(axis=1).iloc[:, -1]).reset_index()    Query  Value0  Value1    PctChg 0    q1     111     222  1.000000 1    q2     333     444  0.333333 

Or

df = pd.DataFrame(td).T.rename_axis('Query').add_prefix('Value') df.eval('PctChg = Value1 / Value0 - 1', inplace=False).reset_index()    Query  Value0  Value1    PctChg 0    q1     111     222  1.000000 1    q2     333     444  0.333333 

Answer 2

Try this ?

td = {'q1':(111,222), 'q2':(333,444)} df = pd.DataFrame(td).T df Out[25]:        0    1 q1  111  222 q2  333  444