#!/usr/bin/env python # coding: utf-8 # In[40]: import h2o # In[41]: h2o.init() # In[42]: from h2o.utils.shared_utils import _locate # private function. used to find files within h2o git project directory. air = h2o.upload_file(_locate("smalldata/airlines/allyears2k_headers.zip")) # In[43]: air.dim # In[44]: numNAs = air["DepTime"].isna().sum() print(numNAs) # In[45]: DepTime_mean = air["DepTime"].mean(na_rm=True) print(DepTime_mean) # In[46]: air.impute("DepTime", method = "median", combine_method="low") numNAs = air["DepTime"].isna().sum() print(numNAs) # In[47]: air = h2o.upload_file(_locate("smalldata/airlines/allyears2k_headers.zip")) # In[48]: air.impute("DepTime", method = "mean", by = ["Origin", "Distance"]) # In[49]: air = h2o.upload_file(_locate("smalldata/airlines/allyears2k_headers.zip")) # In[50]: air.impute("TailNum", method = "mode") # In[51]: air = h2o.upload_file(_locate("smalldata/airlines/allyears2k_headers.zip")) # In[52]: air.impute("TailNum", method = "mode", by=["Month", "Year"])