Raise more helpful exception when concat() wrongly called as concat(obj1,obj2)

[smcinerney]

concat is supposed to be called with objs = a list or dict. Not multiple objects as as separate args. RIGHT: concat([df1,df2]) WRONG: concat(df1,df2)
The exception thrown in the latter case is seriously cryptic.

---

df1 = pd.DataFrame(randn(5,3), columns=['a','b','c'])
df2 = pd.DataFrame(randn(5,3), columns=['a','b','c'])
pd.concat(df1,df2)
... File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/pandas/core/frame.py", line 587, in nonzero
raise ValueError("Cannot call bool() on DataFrame.")

ValueError: Cannot call bool() on DataFrame.

Version: 0.10.1

[jreback]

yep...should fail fast, easy to raise if the objs is not a list or dict.....PR to fix?

[cpcloud]

i gotchu. also allows tuples of things.

merged PR.