Pandas: Values of a data frame doesn't return the expected numpy array.

[drorata]

Code Sample, a copy-pastable example if possible

np.random.seed(42)
arr = np.random.choice(['foo', 'bar', 42], size=(3,3))
df = pd.DataFrame(arr)
print(arr)
print(df)
print(hashlib.sha256(arr.tobytes()).hexdigest())
print(hashlib.sha256(df.values.tobytes()).hexdigest())

Problem description

The hashing suggests that df.values != arr.
Further investigation shows that indeed the types are different.
Moreover, each evaluation of this code yields a new hash for the data frame.

Expected Output

It is expected that pd.DataFrame(np.arr).values == arr.

Output of pd.show_versions()

INSTALLED VERSIONS

commit: None
python: 3.6.1.final.0
python-bits: 64
OS: Darwin
OS-release: 16.5.0
machine: x86_64
processor: i386
byteorder: little
LC_ALL: en_US.UTF-8
LANG: en_US.UTF-8
LOCALE: en_US.UTF-8

pandas: 0.20.1
pytest: 3.0.7
pip: 9.0.1
setuptools: 27.2.0
Cython: 0.25.2
numpy: 1.12.1
scipy: 0.19.0
xarray: None
IPython: 6.0.0
sphinx: 1.5.4
patsy: 0.4.1
dateutil: 2.6.0
pytz: 2017.2
blosc: None
bottleneck: 1.2.0
tables: 3.3.0
numexpr: 2.6.2
feather: None
matplotlib: 2.0.2
openpyxl: 2.4.1
xlrd: 1.0.0
xlwt: 1.2.0
xlsxwriter: 0.9.6
lxml: 3.7.3
bs4: 4.6.0
html5lib: 0.999
sqlalchemy: 1.1.9
pymysql: None
psycopg2: None
jinja2: 2.9.6
s3fs: None
pandas_gbq: None
pandas_datareader: None

[jreback]

.values constructs a consolidated (single) dtyped np.array. since you have object dtypes (strings), this is object. It is newly constructed each time. So hashing doesn't work

In [3]: id(df.values)
Out[3]: 4545762848

In [4]: id(df.values)
Out[4]: 4546427680

However, you can as of 0.20.1, use the included hashing functions which are public (though minimal documentation, except for doc-string), to efficiently hash data. These are a pure data hash and are based on siphashing with a common scheme.

In [5]: from pandas.util import hash_pandas_object

In [6]: hash_pandas_object(df)
Out[6]: 
0     9162640643739096251
1    10885429402166970872
2    13102355359759172147
dtype: uint64

In [7]: hash_pandas_object(df)
Out[7]: 
0     9162640643739096251
1    10885429402166970872
2    13102355359759172147
dtype: uint64