docs: add a code sample for creating a kmeans model

samples/snippets/create_kmeans_model_test.py

@@ -13,20 +13,26 @@
 # limitations under the License.
 
 def test_kmeans_sample():
+<<<<<<< HEAD
 # [START bigquery_dataframes_bqml_kmeans]
 
  import bigframes.pandas as bpd
  import bigframes
  from bigframes import dataframe
+=======
+>>>>>>> 73d2a4681212c1881366c298f44f228bbf208932
  import datetime
 
- #NOTE: ask about line below and whether it is needed outside of colab notebooks
- #bigframes.options.bigquery.project= "username-testing"
+ import bigframes
+ from bigframes import dataframe
+ import bigframes.pandas as bpd
 
+ # NOTE: ask about line below and whether it is needed outside of colab notebooks
+ # bigframes.options.bigquery.project= "username-testing"
  # read_gbq: Loads a DataFrame from BigQuery
  h = bpd.read_gbq("bigquery-public-data.london_bicycles.cycle_hire")
- s= bpd.read_gbq(
- '''
+ s = bpd.read_gbq(
+ """
  SELECT
  id,
  ST_DISTANCE(
@@ -35,43 +41,61 @@ def test_kmeans_sample():
  ) / 1000 AS distance_from_city_center
  FROM
  `bigquery-public-data.london_bicycles.cycle_stations` s
- '''
+ """
  )
  # transform the data
- h= h.rename(columns={"start_station_name": "station_name","start_station_id": "station_id"} )
- h= h[["start_date", "station_name", "station_id", "duration"]]
+ h = h.rename(
+ columns={"start_station_name": "station_name", "start_station_id": "station_id"}
+ )
+ h = h[["start_date", "station_name", "station_id", "duration"]]
 
  # NOTE: line below is not accessed, is it needed outside of colab notebook?
  start_date = datetime.datetime.now()
- sample_time = datetime.datetime(2015, 1, 1, 0, 0, 0, tzinfo= datetime.timezone.utc)
- sample_time2 = datetime.datetime(2016, 1, 1, 0, 0, 0, tzinfo= datetime.timezone.utc)
+ sample_time = datetime.datetime(2015, 1, 1, 0, 0, 0, tzinfo=datetime.timezone.utc)
+ sample_time2 = datetime.datetime(2016, 1, 1, 0, 0, 0, tzinfo=datetime.timezone.utc)
 
- h= h.loc[(h["start_date"] >= sample_time) & (h["start_date"] <= sample_time2)]
+ h = h.loc[(h["start_date"] >= sample_time) & (h["start_date"] <= sample_time2)]
 
- isweekday = h.start_date.dt.dayofweek.map({0: "weekday", 1: "weekday", 2: "weekday", 3: "weekday",
- 4:"weekday",5:"weekend", 6:"weekend"})
+ isweekday = h.start_date.dt.dayofweek.map(
+ {
+ 0: "weekday",
+ 1: "weekday",
+ 2: "weekday",
+ 3: "weekday",
+ 4: "weekday",
+ 5: "weekend",
+ 6: "weekend",
+ }
+ )
 
  # create the new dataframe variable, stationstats
  merged_df = h.merge(
- right= s,
- how="inner",
- left_on= "station_id",
- right_on= "id",
- )
+ right=s,
+ how="inner",
+ left_on="station_id",
+ right_on="id",
+ )
+
+ stationstats = merged_df.groupby("station_name").agg(
+ {"duration": ["mean", "count"], "distance_from_city_center": "max"}
+ )
 
- stationstats = merged_df.groupby("station_name").agg({"duration":[ "mean","count"] , "distance_from_city_center": "max"})
  def station_filter():
- stationstats.columns = ["duration","num_trips","distance_from_city_center"]
- stationstats.sort_values(by = "distance_from_city_center", ascending = True)
- filter = '''REGEXP_CONTAINS(station_name, 'Kennington')'''
+ stationstats.columns = ["duration", "num_trips", "distance_from_city_center"]
+ stationstats.sort_values(by="distance_from_city_center", ascending=True)
+ filter = """REGEXP_CONTAINS(station_name, 'Kennington')"""
 
  # import the KMeans model to cluster the data
  from bigframes.ml.cluster import KMeans
+
  cluster_model = KMeans(n_clusters=4)
- cluster_model= cluster_model.fit(stationstats)
+ cluster_model = cluster_model.fit(stationstats)
 
- #the following function predicts the cluster of every station that has the string "Kennington" in its name.
+ # the following function predicts the cluster of every station that has the string "Kennington" in its name.
  def predict_kennington_stations():
  cluster_model.predict(stationstats)
+<<<<<<< HEAD
 
-# [END bigquery_dataframes_bqml_kmeans]
+# [END bigquery_dataframes_bqml_kmeans]
+=======
+>>>>>>> 73d2a4681212c1881366c298f44f228bbf208932