nathanepstein / dora Goto Github PK

Hi,
I guess I'm missing some very simple detail here...

cat /etc/issue
Ubuntu 16.04 LTS

python3 --version
Python 3.5.1+

pip3 --version
pip 8.1.2 from /usr/local/lib/python3.5/dist-packages/pip-8.1.2-py3.5.egg (python 3.5)

pip3 install Dora
Collecting Dora
  Could not find a version that satisfies the requirement Dora (from versions: )
No matching distribution found for Dora

Here's an MWE

a = pd.DataFrame({"a":[1,2,3]})

dora = Dora(output=0, data=a)

dora.explore()

I believe there is a version error. Upon looking at the setup.py file I noticed that version 0.0.3 was removed and 0.0.4 was added. I tried to install using 0.0.4 but said that the version does not exist.

This all derived from me trying to mitigate the issue of sklearn being deprecated.

Getting this error when i run
dataframe = pd.read_csv("test.csv");
dora = Dora(output = 'A', data = dataframe)
dora.data
dora.impute_missing_values()

features = self.input_data().columns

should be replaced by:
features = self.input_columns()

  def explore(self):
    features = self.input_columns()
    row_count = math.floor(math.sqrt(len(features)))
    col_count = math.ceil(len(features) / row_count)
    figure = plt.figure(1)

Is it suport python2.* ?
thanks

Today I purged my pip cache and reinstalled some packages. So, I saw that Dora uses sklearn and now sklearn gives a message about its own deprecation avoiding the wheel caching mechanism. The recommendation is a change to scikit-learn package.

When I try running the Dora.explore() method, I get an Attribute error:
AttributeError: 'Dora' object has no attribute 'input_data'

I believe this might be because the explore() method in main calls Dora.input_data().columns instead of Dora.input_columns(). Am I missing something/is this fixable? Thanks.