Integration with other tools/frameworks

[mglowacki100]

trafficstars

It'd be great to have a list of recommended tools and frameworks that could be easily integrated with fklearn to have production ready and reproducible pipeline. Even better to have short examples of usage. For example fklearn could be used with: https://pair-code.github.io/what-if-tool/ and mlflow. How about kedro (https://github.com/quantumblacklabs/kedro), do you recommend it?

[caique-lima]

@mglowacki100

Sounds as a nice idea. At nubank, our internal framework is similar to Kedro, so I think we can try to write something related to how they can integrate. I'm unsure if we can have an actual list, but maybe describe how to deploy using one frame can give us a nice idea of how to integrate with other ones

[mglowacki100]

@caique-lima I think toy-example, how to integrate with kedro would be enough, but if you have better approach it also would be great. I'm really close to integrate fklearn with mlflow, but I've issue with mlflow itself. In the meanwhile, here is a sample integration with https://pair-code.github.io/what-if-tool/:

# In[15]:
import dill as pickle
import pandas as pd
from witwidget.notebook.visualization import WitWidget, WitConfigBuilder
# In[16]:
data = pd.read_csv('outime_outspace_hdout.csv')
# In[17]:
data.columns, data.shape
# In[18]:
with open('fkl_model.pkl', 'rb') as pickle_file:
    prediction_function = pickle.load(pickle_file)
# In[19]:
config_builder = (WitConfigBuilder(data.values.tolist(), data.columns.tolist())
  .set_custom_predict_fn(prediction_function)
  .set_target_feature('y')
  .set_model_type('regression'))
# In[20]:
WitWidget(config_builder, height=800)