Formula 1 - Predicting the winner of a race¶

In this notebook we will benchmark getML against featuretools to predict the winner of a Formula 1 race.

Summary:

Prediction type: Classification model
Domain: Sports
Prediction target: Win
Population size: 31578

Background¶

We would like to develop a prediction model for Formula 1 races, that would allow us to predict the winner of a race before the race has started.

We use dataset of all Formula 1 races from 1950 to 2017. The dataset includes information such as the time taken in each lap, the time taken for pit stops, the performance in the qualifying rounds etc.

The dataset has been downloaded from the CTU Prague relational learning repository (Motl and Schulte, 2015) (Now residing at relational-data.org.).

We will benchmark getML's feature learning algorithms against featuretools, an open-source implementation of the propositionalization algorithm, similar to getML's FastProp.

Analysis¶

In [1]:

  Copied!     
 
%pip install -q "getml==1.5.0" "matplotlib==3.9.2" "featuretools==1.31.0" "ipywidgets==8.1.5"
%pip install -q "getml==1.5.0" "matplotlib==3.9.2" "featuretools==1.31.0" "ipywidgets==8.1.5"

Let's get started with the analysis and set up your session:

In [2]:

  Copied!     
 
import os
import warnings

import pandas as pd
import matplotlib.pyplot as plt

import featuretools
import woodwork as ww
import getml

os.environ["PYARROW_IGNORE_TIMEZONE"] = "1"
warnings.simplefilter(action='ignore', category=FutureWarning)
%matplotlib inline

print(f"getML API version: {getml.__version__}\n")
import os import warnings import pandas as pd import matplotlib.pyplot as plt import featuretools import woodwork as ww import getml os.environ["PYARROW_IGNORE_TIMEZONE"] = "1" warnings.simplefilter(action='ignore', category=FutureWarning) %matplotlib inline print(f"getML API version: {getml.__version__}\n")

getML API version: 1.5.0

In [59]:

  Copied!     
 
getml.engine.launch(allow_remote_ips=True, token='token')
getml.engine.set_project('formula1')
getml.engine.launch(allow_remote_ips=True, token='token') getml.engine.set_project('formula1')

Launching ./getML --allow-push-notifications=true --allow-remote-ips=true --home-directory=/home/user --in-memory=true --install=false --launch-browser=true --log=false --token=token in /home/user/.getML/getml-1.5.0-x64-linux...
Launched the getML Engine. The log output will be stored in /home/user/.getML/logs/20240912140548.log.
  Loading pipelines... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00

Connected to project 'formula1'.

1. Loading data¶

1.1 Download from source¶

We begin by downloading the data:

In [4]:

  Copied!     
 
conn = getml.database.connect_mysql(
    host="db.relational-data.org",
    dbname="ErgastF1",
    port=3306,
    user="guest",
    password="relational"
)

conn
conn = getml.database.connect_mysql( host="db.relational-data.org", dbname="ErgastF1", port=3306, user="guest", password="relational" ) conn

Out[4]:

Connection(dbname='ErgastF1',
           dialect='mysql',
           host='db.relational-data.org',
           port=3306)

In [5]:

  Copied!     
 
def load_if_needed(name):
    """
    Loads the data from the relational learning
    repository, if the data frame has not already
    been loaded.
    """
    if not getml.data.exists(name):
        data_frame = getml.data.DataFrame.from_db(
            name=name,
            table_name=name,
            conn=conn
        )
        data_frame.save()
    else:
        data_frame = getml.data.load_data_frame(name)
    return data_frame
def load_if_needed(name): """ Loads the data from the relational learning repository, if the data frame has not already been loaded. """ if not getml.data.exists(name): data_frame = getml.data.DataFrame.from_db( name=name, table_name=name, conn=conn ) data_frame.save() else: data_frame = getml.data.load_data_frame(name) return data_frame

In [6]:

  Copied!     
 
driverStandings = load_if_needed("driverStandings")
drivers = load_if_needed("drivers")
lapTimes = load_if_needed("lapTimes")
pitStops = load_if_needed("pitStops")
races = load_if_needed("races")
qualifying = load_if_needed("qualifying")
driverStandings = load_if_needed("driverStandings") drivers = load_if_needed("drivers") lapTimes = load_if_needed("lapTimes") pitStops = load_if_needed("pitStops") races = load_if_needed("races") qualifying = load_if_needed("qualifying")

In [7]:

  Copied!     
 
driverStandings
driverStandings

Out[7]:

name	driverStandingsId	raceId	driverId	points	position	wins	positionText
role	unused_float	unused_float	unused_float	unused_float	unused_float	unused_float	unused_string
0	1	18	1	10	1	1	1
1	2	18	2	8	2	0	2
2	3	18	3	6	3	0	3
3	4	18	4	5	4	0	4
4	5	18	5	4	5	0	5
	...	...	...	...	...	...	...
31573	68456	982	835	8	16	0	16
31574	68457	982	154	26	13	0	13
31575	68458	982	836	5	18	0	18
31576	68459	982	18	0	22	0	22
31577	68460	982	814	0	23	0	23

31578 rows x 7 columns
memory usage: 1.85 MB
name: driverStandings
type: getml.DataFrame

In [8]:

  Copied!     
 
drivers
drivers

Out[8]:

name	driverId	number	driverRef	code	forename	surname	dob	nationality	url
role	unused_float	unused_float	unused_string	unused_string	unused_string	unused_string	unused_string	unused_string	unused_string
0	1	44	hamilton	HAM	Lewis	Hamilton	1985-01-07	British	http://en.wikipedia.org/wiki/Lew...
1	2	nan	heidfeld	HEI	Nick	Heidfeld	1977-05-10	German	http://en.wikipedia.org/wiki/Nic...
2	3	6	rosberg	ROS	Nico	Rosberg	1985-06-27	German	http://en.wikipedia.org/wiki/Nic...
3	4	14	alonso	ALO	Fernando	Alonso	1981-07-29	Spanish	http://en.wikipedia.org/wiki/Fer...
4	5	nan	kovalainen	KOV	Heikki	Kovalainen	1981-10-19	Finnish	http://en.wikipedia.org/wiki/Hei...
	...	...	...	...	...	...	...	...	...
835	837	88	haryanto	HAR	Rio	Haryanto	1993-01-22	Indonesian	http://en.wikipedia.org/wiki/Rio...
836	838	2	vandoorne	VAN	Stoffel	Vandoorne	1992-03-26	Belgian	http://en.wikipedia.org/wiki/Sto...
837	839	31	ocon	OCO	Esteban	Ocon	1996-09-17	French	http://en.wikipedia.org/wiki/Est...
838	840	18	stroll	STR	Lance	Stroll	1998-10-29	Canadian	http://en.wikipedia.org/wiki/Lan...
839	841	36	giovinazzi	GIO	Antonio	Giovinazzi	1993-12-14	Italian	http://en.wikipedia.org/wiki/Ant...

840 rows x 9 columns
memory usage: 0.13 MB
name: drivers
type: getml.DataFrame

In [9]:

  Copied!     
 
lapTimes
lapTimes

Out[9]:

name	raceId	driverId	lap	position	milliseconds	time
role	unused_float	unused_float	unused_float	unused_float	unused_float	unused_string
0	1	1	1	13	109088	1:49.088
1	1	1	2	12	93740	1:33.740
2	1	1	3	11	91600	1:31.600
3	1	1	4	10	91067	1:31.067
4	1	1	5	10	92129	1:32.129
	...	...	...	...	...	...
420364	982	840	54	8	107528	1:47.528
420365	982	840	55	8	107512	1:47.512
420366	982	840	56	8	108143	1:48.143
420367	982	840	57	8	107848	1:47.848
420368	982	840	58	8	108699	1:48.699

420369 rows x 6 columns
memory usage: 23.96 MB
name: lapTimes
type: getml.DataFrame

In [10]:

  Copied!     
 
pitStops
pitStops

Out[10]:

name	raceId	driverId	stop	lap	milliseconds	time	duration
role	unused_float	unused_float	unused_float	unused_float	unused_float	unused_string	unused_string
0	841	1	1	16	23227	17:28:24	23.227
1	841	1	2	36	23199	17:59:29	23.199
2	841	2	1	15	22994	17:27:41	22.994
3	841	2	2	30	25098	17:51:32	25.098
4	841	3	1	16	23716	17:29:00	23.716
	...	...	...	...	...	...	...
6065	982	839	6	38	29134	21:29:07	29.134
6066	982	840	1	1	37403	20:06:43	37.403
6067	982	840	2	2	29294	20:10:07	29.294
6068	982	840	3	3	25584	20:13:16	25.584
6069	982	840	4	26	29412	21:05:07	29.412

6070 rows x 7 columns
memory usage: 0.44 MB
name: pitStops
type: getml.DataFrame

In [11]:

  Copied!     
 
races
races

Out[11]:

name	raceId	year	round	circuitId	name	date	time	url
role	unused_float	unused_float	unused_float	unused_float	unused_string	unused_string	unused_string	unused_string
0	1	2009	1	1	Australian Grand Prix	2009-03-29	06:00:00	http://en.wikipedia.org/wiki/200...
1	2	2009	2	2	Malaysian Grand Prix	2009-04-05	09:00:00	http://en.wikipedia.org/wiki/200...
2	3	2009	3	17	Chinese Grand Prix	2009-04-19	07:00:00	http://en.wikipedia.org/wiki/200...
3	4	2009	4	3	Bahrain Grand Prix	2009-04-26	12:00:00	http://en.wikipedia.org/wiki/200...
4	5	2009	5	4	Spanish Grand Prix	2009-05-10	12:00:00	http://en.wikipedia.org/wiki/200...
	...	...	...	...	...	...	...	...
971	984	2017	16	22	Japanese Grand Prix	2017-10-08	05:00:00	https://en.wikipedia.org/wiki/20...
972	985	2017	17	69	United States Grand Prix	2017-10-22	19:00:00	https://en.wikipedia.org/wiki/20...
973	986	2017	18	32	Mexican Grand Prix	2017-10-29	19:00:00	https://en.wikipedia.org/wiki/20...
974	987	2017	19	18	Brazilian Grand Prix	2017-11-12	16:00:00	https://en.wikipedia.org/wiki/20...
975	988	2017	20	24	Abu Dhabi Grand Prix	2017-11-26	17:00:00	https://en.wikipedia.org/wiki/20...

976 rows x 8 columns
memory usage: 0.15 MB
name: races
type: getml.DataFrame

In [12]:

  Copied!     
 
qualifying
qualifying

Out[12]:

name	qualifyId	raceId	driverId	constructorId	number	position	q1	q2	q3
role	unused_float	unused_float	unused_float	unused_float	unused_float	unused_float	unused_string	unused_string	unused_string
0	1	18	1	1	22	1	1:26.572	1:25.187	1:26.714
1	2	18	9	2	4	2	1:26.103	1:25.315	1:26.869
2	3	18	5	1	23	3	1:25.664	1:25.452	1:27.079
3	4	18	13	6	2	4	1:25.994	1:25.691	1:27.178
4	5	18	2	2	3	5	1:25.960	1:25.518	1:27.236
	...	...	...	...	...	...	...	...	...
7392	7415	982	825	210	20	16	1:43.756	NULL	NULL
7393	7416	982	13	3	19	17	1:44.014	NULL	NULL
7394	7417	982	840	3	18	18	1:44.728	NULL	NULL
7395	7418	982	836	15	94	19	1:45.059	NULL	NULL
7396	7419	982	828	15	9	20	1:45.570	NULL	NULL

7397 rows x 9 columns
memory usage: 0.66 MB
name: qualifying
type: getml.DataFrame

1.2 Prepare data for getML¶

In [13]:

  Copied!     
 
racesPd = races.to_pandas()
racesPd
racesPd = races.to_pandas() racesPd

Out[13]:

	raceId	year	round	circuitId	name	date	time	url
0	1.0	2009.0	1.0	1.0	Australian Grand Prix	2009-03-29	06:00:00	http://en.wikipedia.org/wiki/2009_Australian_G...
1	2.0	2009.0	2.0	2.0	Malaysian Grand Prix	2009-04-05	09:00:00	http://en.wikipedia.org/wiki/2009_Malaysian_Gr...
2	3.0	2009.0	3.0	17.0	Chinese Grand Prix	2009-04-19	07:00:00	http://en.wikipedia.org/wiki/2009_Chinese_Gran...
3	4.0	2009.0	4.0	3.0	Bahrain Grand Prix	2009-04-26	12:00:00	http://en.wikipedia.org/wiki/2009_Bahrain_Gran...
4	5.0	2009.0	5.0	4.0	Spanish Grand Prix	2009-05-10	12:00:00	http://en.wikipedia.org/wiki/2009_Spanish_Gran...
...	...	...	...	...	...	...	...	...
971	984.0	2017.0	16.0	22.0	Japanese Grand Prix	2017-10-08	05:00:00	https://en.wikipedia.org/wiki/2017_Japanese_Gr...
972	985.0	2017.0	17.0	69.0	United States Grand Prix	2017-10-22	19:00:00	https://en.wikipedia.org/wiki/2017_United_Stat...
973	986.0	2017.0	18.0	32.0	Mexican Grand Prix	2017-10-29	19:00:00	https://en.wikipedia.org/wiki/2017_Mexican_Gra...
974	987.0	2017.0	19.0	18.0	Brazilian Grand Prix	2017-11-12	16:00:00	https://en.wikipedia.org/wiki/2017_Brazilian_G...
975	988.0	2017.0	20.0	24.0	Abu Dhabi Grand Prix	2017-11-26	17:00:00	https://en.wikipedia.org/wiki/2017_Abu_Dhabi_G...

976 rows × 8 columns

We actually need some set-up, because the target variable is not readily available. The wins column in driverStandings is actually the accumulated number of wins over a year, but what we want is a boolean variable indicated whether someone has one a particular race or not.

In [14]:

  Copied!     
 
driverStandingsPd = driverStandings.to_pandas()

driverStandingsPd = driverStandingsPd.merge(
    racesPd[["raceId", "year", "date", "round"]],
    on="raceId"
)

previousStanding = driverStandingsPd.merge(
    driverStandingsPd[["driverId", "year", "wins", "round"]],
    on=["driverId", "year"],
)

isPreviousRound = (previousStanding["round_x"] - previousStanding["round_y"] == 1.0)

previousStanding = previousStanding[isPreviousRound]

previousStanding["win"] = previousStanding["wins_x"] - previousStanding["wins_y"]

driverStandingsPd = driverStandingsPd.merge(
    previousStanding[["raceId", "driverId", "win"]],
    on=["raceId", "driverId"],
    how="left",
)

driverStandingsPd["win"] = [win if win == win else wins for win, wins in zip(driverStandingsPd["win"], driverStandingsPd["wins"])]

driver_standings = getml.data.DataFrame.from_pandas(driverStandingsPd, "driver_standings")

driver_standings
driverStandingsPd = driverStandings.to_pandas() driverStandingsPd = driverStandingsPd.merge( racesPd[["raceId", "year", "date", "round"]], on="raceId" ) previousStanding = driverStandingsPd.merge( driverStandingsPd[["driverId", "year", "wins", "round"]], on=["driverId", "year"], ) isPreviousRound = (previousStanding["round_x"] - previousStanding["round_y"] == 1.0) previousStanding = previousStanding[isPreviousRound] previousStanding["win"] = previousStanding["wins_x"] - previousStanding["wins_y"] driverStandingsPd = driverStandingsPd.merge( previousStanding[["raceId", "driverId", "win"]], on=["raceId", "driverId"], how="left", ) driverStandingsPd["win"] = [win if win == win else wins for win, wins in zip(driverStandingsPd["win"], driverStandingsPd["wins"])] driver_standings = getml.data.DataFrame.from_pandas(driverStandingsPd, "driver_standings") driver_standings

Out[14]:

name	driverStandingsId	raceId	driverId	points	position	wins	year	round	win	positionText	date
role	unused_float	unused_float	unused_float	unused_float	unused_float	unused_float	unused_float	unused_float	unused_float	unused_string	unused_string
0	1	18	1	10	1	1	2008	1	1	1	2008-03-16
1	2	18	2	8	2	0	2008	1	0	2	2008-03-16
2	3	18	3	6	3	0	2008	1	0	3	2008-03-16
3	4	18	4	5	4	0	2008	1	0	4	2008-03-16
4	5	18	5	4	5	0	2008	1	0	5	2008-03-16
	...	...	...	...	...	...	...	...	...	...	...
31573	68456	982	835	8	16	0	2017	14	0	16	2017-09-17
31574	68457	982	154	26	13	0	2017	14	0	13	2017-09-17
31575	68458	982	836	5	18	0	2017	14	0	18	2017-09-17
31576	68459	982	18	0	22	0	2017	14	0	22	2017-09-17
31577	68460	982	814	0	23	0	2017	14	0	23	2017-09-17

31578 rows x 11 columns
memory usage: 3.21 MB
name: driver_standings
type: getml.DataFrame

We also need to include the date of the race to lapTimes and pitStops, because we cannot use this data for the race we would like to predict. We can only take lap times and pit stops from previous races.

In [15]:

  Copied!     
 
lapTimesPd = lapTimes.to_pandas()

lapTimesPd = lapTimesPd.merge(
    racesPd[["raceId", "date", "year"]],
    on="raceId"
)

lap_times = getml.data.DataFrame.from_pandas(lapTimesPd, "lap_times")

lap_times
lapTimesPd = lapTimes.to_pandas() lapTimesPd = lapTimesPd.merge( racesPd[["raceId", "date", "year"]], on="raceId" ) lap_times = getml.data.DataFrame.from_pandas(lapTimesPd, "lap_times") lap_times

Out[15]:

name	raceId	driverId	lap	position	milliseconds	year	time	date
role	unused_float	unused_float	unused_float	unused_float	unused_float	unused_float	unused_string	unused_string
0	1	1	1	13	109088	2009	1:49.088	2009-03-29
1	1	1	2	12	93740	2009	1:33.740	2009-03-29
2	1	1	3	11	91600	2009	1:31.600	2009-03-29
3	1	1	4	10	91067	2009	1:31.067	2009-03-29
4	1	1	5	10	92129	2009	1:32.129	2009-03-29
	...	...	...	...	...	...	...	...
420364	982	840	54	8	107528	2017	1:47.528	2017-09-17
420365	982	840	55	8	107512	2017	1:47.512	2017-09-17
420366	982	840	56	8	108143	2017	1:48.143	2017-09-17
420367	982	840	57	8	107848	2017	1:47.848	2017-09-17
420368	982	840	58	8	108699	2017	1:48.699	2017-09-17

420369 rows x 8 columns
memory usage: 35.31 MB
name: lap_times
type: getml.DataFrame

In [16]:

  Copied!     
 
pitStopsPd = pitStops.to_pandas()

pitStopsPd = pitStopsPd.merge(
    racesPd[["raceId", "date", "year"]],
    on="raceId"
)

pit_stops = getml.data.DataFrame.from_pandas(pitStopsPd, "pit_stops")

pit_stops
pitStopsPd = pitStops.to_pandas() pitStopsPd = pitStopsPd.merge( racesPd[["raceId", "date", "year"]], on="raceId" ) pit_stops = getml.data.DataFrame.from_pandas(pitStopsPd, "pit_stops") pit_stops

Out[16]:

name	raceId	driverId	stop	lap	milliseconds	year	time	duration	date
role	unused_float	unused_float	unused_float	unused_float	unused_float	unused_float	unused_string	unused_string	unused_string
0	841	1	1	16	23227	2011	17:28:24	23.227	2011-03-27
1	841	1	2	36	23199	2011	17:59:29	23.199	2011-03-27
2	841	2	1	15	22994	2011	17:27:41	22.994	2011-03-27
3	841	2	2	30	25098	2011	17:51:32	25.098	2011-03-27
4	841	3	1	16	23716	2011	17:29:00	23.716	2011-03-27
	...	...	...	...	...	...	...	...	...
6065	982	839	6	38	29134	2017	21:29:07	29.134	2017-09-17
6066	982	840	1	1	37403	2017	20:06:43	37.403	2017-09-17
6067	982	840	2	2	29294	2017	20:10:07	29.294	2017-09-17
6068	982	840	3	3	25584	2017	20:13:16	25.584	2017-09-17
6069	982	840	4	26	29412	2017	21:05:07	29.412	2017-09-17

6070 rows x 9 columns
memory usage: 0.60 MB
name: pit_stops
type: getml.DataFrame

getML requires that we define roles for each of the columns.

In [17]:

  Copied!     
 
driver_standings.set_role("win", getml.data.roles.target)
driver_standings.set_role(["raceId", "driverId", "year"], getml.data.roles.join_key)
driver_standings.set_role("position", getml.data.roles.numerical)
driver_standings.set_role("date", getml.data.roles.time_stamp)

driver_standings
driver_standings.set_role("win", getml.data.roles.target) driver_standings.set_role(["raceId", "driverId", "year"], getml.data.roles.join_key) driver_standings.set_role("position", getml.data.roles.numerical) driver_standings.set_role("date", getml.data.roles.time_stamp) driver_standings

Out[17]:

name	date	raceId	driverId	year	win	position	driverStandingsId	points	wins	round	positionText
role	time_stamp	join_key	join_key	join_key	target	numerical	unused_float	unused_float	unused_float	unused_float	unused_string
unit	time stamp, comparison only
0	2008-03-16	18	1	2008	1	1	1	10	1	1	1
1	2008-03-16	18	2	2008	0	2	2	8	0	1	2
2	2008-03-16	18	3	2008	0	3	3	6	0	1	3
3	2008-03-16	18	4	2008	0	4	4	5	0	1	4
4	2008-03-16	18	5	2008	0	5	5	4	0	1	5
	...	...	...	...	...	...	...	...	...	...	...
31573	2017-09-17	982	835	2017	0	16	68456	8	0	14	16
31574	2017-09-17	982	154	2017	0	13	68457	26	0	14	13
31575	2017-09-17	982	836	2017	0	18	68458	5	0	14	18
31576	2017-09-17	982	18	2017	0	22	68459	0	0	14	22
31577	2017-09-17	982	814	2017	0	23	68460	0	0	14	23

31578 rows x 11 columns
memory usage: 2.49 MB
name: driver_standings
type: getml.DataFrame

In [18]:

  Copied!     
 
drivers.set_role("driverId", getml.data.roles.join_key)
drivers.set_role(["nationality", "driverRef"], getml.data.roles.categorical)

drivers
drivers.set_role("driverId", getml.data.roles.join_key) drivers.set_role(["nationality", "driverRef"], getml.data.roles.categorical) drivers

Out[18]:

name	driverId	nationality	driverRef	number	code	forename	surname	dob	url
role	join_key	categorical	categorical	unused_float	unused_string	unused_string	unused_string	unused_string	unused_string
0	1	British	hamilton	44	HAM	Lewis	Hamilton	1985-01-07	http://en.wikipedia.org/wiki/Lew...
1	2	German	heidfeld	nan	HEI	Nick	Heidfeld	1977-05-10	http://en.wikipedia.org/wiki/Nic...
2	3	German	rosberg	6	ROS	Nico	Rosberg	1985-06-27	http://en.wikipedia.org/wiki/Nic...
3	4	Spanish	alonso	14	ALO	Fernando	Alonso	1981-07-29	http://en.wikipedia.org/wiki/Fer...
4	5	Finnish	kovalainen	nan	KOV	Heikki	Kovalainen	1981-10-19	http://en.wikipedia.org/wiki/Hei...
	...	...	...	...	...	...	...	...	...
835	837	Indonesian	haryanto	88	HAR	Rio	Haryanto	1993-01-22	http://en.wikipedia.org/wiki/Rio...
836	838	Belgian	vandoorne	2	VAN	Stoffel	Vandoorne	1992-03-26	http://en.wikipedia.org/wiki/Sto...
837	839	French	ocon	31	OCO	Esteban	Ocon	1996-09-17	http://en.wikipedia.org/wiki/Est...
838	840	Canadian	stroll	18	STR	Lance	Stroll	1998-10-29	http://en.wikipedia.org/wiki/Lan...
839	841	Italian	giovinazzi	36	GIO	Antonio	Giovinazzi	1993-12-14	http://en.wikipedia.org/wiki/Ant...

840 rows x 9 columns
memory usage: 0.11 MB
name: drivers
type: getml.DataFrame

In [19]:

  Copied!     
 
lap_times.set_role(["raceId", "driverId", "year"], getml.data.roles.join_key)
lap_times.set_role(["lap", "milliseconds", "position"], getml.data.roles.numerical)
lap_times.set_role("date", getml.data.roles.time_stamp)

lap_times
lap_times.set_role(["raceId", "driverId", "year"], getml.data.roles.join_key) lap_times.set_role(["lap", "milliseconds", "position"], getml.data.roles.numerical) lap_times.set_role("date", getml.data.roles.time_stamp) lap_times

Out[19]:

name	date	raceId	driverId	year	lap	milliseconds	position	time
role	time_stamp	join_key	join_key	join_key	numerical	numerical	numerical	unused_string
unit	time stamp, comparison only
0	2009-03-29	1	1	2009	1	109088	13	1:49.088
1	2009-03-29	1	1	2009	2	93740	12	1:33.740
2	2009-03-29	1	1	2009	3	91600	11	1:31.600
3	2009-03-29	1	1	2009	4	91067	10	1:31.067
4	2009-03-29	1	1	2009	5	92129	10	1:32.129
	...	...	...	...	...	...	...	...
420364	2017-09-17	982	840	2017	54	107528	8	1:47.528
420365	2017-09-17	982	840	2017	55	107512	8	1:47.512
420366	2017-09-17	982	840	2017	56	108143	8	1:48.143
420367	2017-09-17	982	840	2017	57	107848	8	1:47.848
420368	2017-09-17	982	840	2017	58	108699	8	1:48.699

420369 rows x 8 columns
memory usage: 25.64 MB
name: lap_times
type: getml.DataFrame

In [20]:

  Copied!     
 
pit_stops.set_role(["raceId", "driverId", "year"], getml.data.roles.join_key)
pit_stops.set_role(["lap", "milliseconds", "stop"], getml.data.roles.numerical)
pit_stops.set_role("date", getml.data.roles.time_stamp)

pit_stops
pit_stops.set_role(["raceId", "driverId", "year"], getml.data.roles.join_key) pit_stops.set_role(["lap", "milliseconds", "stop"], getml.data.roles.numerical) pit_stops.set_role("date", getml.data.roles.time_stamp) pit_stops

Out[20]:

name	date	raceId	driverId	year	lap	milliseconds	stop	time	duration
role	time_stamp	join_key	join_key	join_key	numerical	numerical	numerical	unused_string	unused_string
unit	time stamp, comparison only
0	2011-03-27	841	1	2011	16	23227	1	17:28:24	23.227
1	2011-03-27	841	1	2011	36	23199	2	17:59:29	23.199
2	2011-03-27	841	2	2011	15	22994	1	17:27:41	22.994
3	2011-03-27	841	2	2011	30	25098	2	17:51:32	25.098
4	2011-03-27	841	3	2011	16	23716	1	17:29:00	23.716
	...	...	...	...	...	...	...	...	...
6065	2017-09-17	982	839	2017	38	29134	6	21:29:07	29.134
6066	2017-09-17	982	840	2017	1	37403	1	20:06:43	37.403
6067	2017-09-17	982	840	2017	2	29294	2	20:10:07	29.294
6068	2017-09-17	982	840	2017	3	25584	3	20:13:16	25.584
6069	2017-09-17	982	840	2017	26	29412	4	21:05:07	29.412

6070 rows x 9 columns
memory usage: 0.46 MB
name: pit_stops
type: getml.DataFrame

In [21]:

  Copied!     
 
qualifying.set_role(["raceId", "driverId", "qualifyId"], getml.data.roles.join_key)
qualifying.set_role(["position", "number"], getml.data.roles.numerical)

qualifying
qualifying.set_role(["raceId", "driverId", "qualifyId"], getml.data.roles.join_key) qualifying.set_role(["position", "number"], getml.data.roles.numerical) qualifying

Out[21]:

name	raceId	driverId	qualifyId	position	number	constructorId	q1	q2	q3
role	join_key	join_key	join_key	numerical	numerical	unused_float	unused_string	unused_string	unused_string
0	18	1	1	1	22	1	1:26.572	1:25.187	1:26.714
1	18	9	2	2	4	2	1:26.103	1:25.315	1:26.869
2	18	5	3	3	23	1	1:25.664	1:25.452	1:27.079
3	18	13	4	4	2	6	1:25.994	1:25.691	1:27.178
4	18	2	5	5	3	2	1:25.960	1:25.518	1:27.236
	...	...	...	...	...	...	...	...	...
7392	982	825	7415	16	20	210	1:43.756	NULL	NULL
7393	982	13	7416	17	19	3	1:44.014	NULL	NULL
7394	982	840	7417	18	18	3	1:44.728	NULL	NULL
7395	982	836	7418	19	94	15	1:45.059	NULL	NULL
7396	982	828	7419	20	9	15	1:45.570	NULL	NULL

7397 rows x 9 columns
memory usage: 0.57 MB
name: qualifying
type: getml.DataFrame

2. Predictive modeling¶

We loaded the data and defined the roles and units. Next, we create a getML pipeline for relational learning.

In [22]:

  Copied!     
 
split = getml.data.split.random(train=0.8, test=0.2)
split
split = getml.data.split.random(train=0.8, test=0.2) split

Out[22]:


0	train
1	train
2	train
3	test
4	train
	...

infinite number of rows
type: StringColumnView

2.1 Define relational model¶

In [23]:

  Copied!     
 
star_schema = getml.data.StarSchema(population=driver_standings.drop(["position"]), alias="population", split=split)

star_schema.join(
    driver_standings,
    on=["driverId"],
    time_stamps="date",
    horizon=getml.data.time.days(1),
    lagged_targets=True,
)

# We cannot use lap times for the race
# we would like to predict, so we set
# a non-zero horizon.
star_schema.join(
    lap_times,
    on=["driverId"],
    time_stamps="date",
    horizon=getml.data.time.days(1),
)

# We cannot use pit stops for the race
# we would like to predict, so we set
# a non-zero horizon.
star_schema.join(
    pit_stops,
    on=["driverId"],
    time_stamps="date",
    horizon=getml.data.time.days(1),
)

star_schema.join(
    qualifying,
    on=["driverId", "raceId"],
    relationship=getml.data.relationship.many_to_one,
)

star_schema.join(
    drivers,
    on=["driverId"],
    relationship=getml.data.relationship.many_to_one,
)

star_schema
star_schema = getml.data.StarSchema(population=driver_standings.drop(["position"]), alias="population", split=split) star_schema.join( driver_standings, on=["driverId"], time_stamps="date", horizon=getml.data.time.days(1), lagged_targets=True, ) # We cannot use lap times for the race # we would like to predict, so we set # a non-zero horizon. star_schema.join( lap_times, on=["driverId"], time_stamps="date", horizon=getml.data.time.days(1), ) # We cannot use pit stops for the race # we would like to predict, so we set # a non-zero horizon. star_schema.join( pit_stops, on=["driverId"], time_stamps="date", horizon=getml.data.time.days(1), ) star_schema.join( qualifying, on=["driverId", "raceId"], relationship=getml.data.relationship.many_to_one, ) star_schema.join( drivers, on=["driverId"], relationship=getml.data.relationship.many_to_one, ) star_schema

Out[23]:

data model

diagram

staging

	data frames	staging table
0	population, qualifying, drivers	POPULATION__STAGING_TABLE_1
1	driver_standings	DRIVER_STANDINGS__STAGING_TABLE_2
2	lap_times	LAP_TIMES__STAGING_TABLE_3
3	pit_stops	PIT_STOPS__STAGING_TABLE_4

container

population

	subset	name	rows	type
0	test	driver_standings	6229	View
1	train	driver_standings	25349	View

peripheral

	name	rows	type
0	driver_standings	31578	DataFrame
1	lap_times	420369	DataFrame
2	pit_stops	6070	DataFrame
3	qualifying	7397	DataFrame
4	drivers	840	DataFrame

2.2 getML pipeline¶

Set-up the feature learner & predictor

We use the relboost algorithms for this problem. Because of the large number of keywords, we regularize the model a bit by requiring a minimum support for the keywords (min_num_samples).

In [24]:

  Copied!     
 
mapping = getml.preprocessors.Mapping()

fast_prop = getml.feature_learning.FastProp(
    loss_function=getml.feature_learning.loss_functions.CrossEntropyLoss,
    aggregation=getml.feature_learning.FastProp.agg_sets.All,
    num_threads=1,
)

relboost = getml.feature_learning.Relboost(
    loss_function=getml.feature_learning.loss_functions.CrossEntropyLoss,
    num_threads=1,
)

relmt = getml.feature_learning.RelMT(
    loss_function=getml.feature_learning.loss_functions.CrossEntropyLoss,
    num_threads=1,
)

predictor = getml.predictors.XGBoostClassifier(n_jobs=1)
mapping = getml.preprocessors.Mapping() fast_prop = getml.feature_learning.FastProp( loss_function=getml.feature_learning.loss_functions.CrossEntropyLoss, aggregation=getml.feature_learning.FastProp.agg_sets.All, num_threads=1, ) relboost = getml.feature_learning.Relboost( loss_function=getml.feature_learning.loss_functions.CrossEntropyLoss, num_threads=1, ) relmt = getml.feature_learning.RelMT( loss_function=getml.feature_learning.loss_functions.CrossEntropyLoss, num_threads=1, ) predictor = getml.predictors.XGBoostClassifier(n_jobs=1)

Build the pipeline

In [25]:

  Copied!     
 
pipe1 = getml.pipeline.Pipeline(
    tags=['fast_prop'],
    data_model=star_schema.data_model,
    preprocessors=[mapping],
    feature_learners=[fast_prop],
    predictors=[predictor],
    include_categorical=True,
)

pipe1
pipe1 = getml.pipeline.Pipeline( tags=['fast_prop'], data_model=star_schema.data_model, preprocessors=[mapping], feature_learners=[fast_prop], predictors=[predictor], include_categorical=True, ) pipe1

Out[25]:

Pipeline(data_model='population',
         feature_learners=['FastProp'],
         feature_selectors=[],
         include_categorical=True,
         loss_function='CrossEntropyLoss',
         peripheral=['driver_standings', 'drivers', 'lap_times', 'pit_stops', 'qualifying'],
         predictors=['XGBoostClassifier'],
         preprocessors=['Mapping'],
         share_selected_features=0.5,
         tags=['fast_prop'])

2.3 Model training¶

In [26]:

  Copied!     
 
pipe1.check(star_schema.train)
pipe1.check(star_schema.train)

Checking data model...

  Staging... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00
  Preprocessing... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:05
  Checking... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00

The pipeline check generated 2 issues labeled INFO and 0 issues labeled WARNING.

Out[26]:

	type	label	message
0	INFO	FOREIGN KEYS NOT FOUND	When joining POPULATION__STAGING_TABLE_1 and LAP_TIMES__STAGING_TABLE_3 over 'driverId' and 'driverId', there are no corresponding entries for 68.551028% of entries in 'driverId' in 'POPULATION__STAGING_TABLE_1'. You might want to double-check your join keys.
1	INFO	FOREIGN KEYS NOT FOUND	When joining POPULATION__STAGING_TABLE_1 and PIT_STOPS__STAGING_TABLE_4 over 'driverId' and 'driverId', there are no corresponding entries for 82.527910% of entries in 'driverId' in 'POPULATION__STAGING_TABLE_1'. You might want to double-check your join keys.

In [27]:

  Copied!     
 
pipe1.fit(star_schema.train)
pipe1.fit(star_schema.train)

Checking data model...

  Staging... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00
  Preprocessing... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00

The pipeline check generated 2 issues labeled INFO and 0 issues labeled WARNING.

To see the issues in full, run .check() on the pipeline.

  Staging... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00
  Preprocessing... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00
  FastProp: Trying 993 features... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 02:35
  FastProp: Building features... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:35
  XGBoost: Training as predictor... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:15

Trained pipeline.

Time taken: 0:03:26.520434.

Out[27]:

Pipeline(data_model='population',
         feature_learners=['FastProp'],
         feature_selectors=[],
         include_categorical=True,
         loss_function='CrossEntropyLoss',
         peripheral=['driver_standings', 'drivers', 'lap_times', 'pit_stops', 'qualifying'],
         predictors=['XGBoostClassifier'],
         preprocessors=['Mapping'],
         share_selected_features=0.5,
         tags=['fast_prop', 'container-DlsHvD'])

2.4 Model evaluation¶

In [28]:

  Copied!     
 
fastprop_score = pipe1.score(star_schema.test)
fastprop_score
fastprop_score = pipe1.score(star_schema.test) fastprop_score

  Staging... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00
  Preprocessing... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00
  FastProp: Building features... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:09

Out[28]:

	date time	set used	target	accuracy	auc	cross entropy
0	2024-09-12 12:53:27	train	win	0.9732	0.9577	0.07529
1	2024-09-12 12:53:37	test	win	0.9724	0.9235	0.08533

2.5 featuretools¶

featuretools requires some additional data preparation before we can start engineering features.

In [29]:

  Copied!     
 
population_train_pd = star_schema.train.population.drop(["position"]).to_pandas()
population_test_pd = star_schema.test.population.drop(["position"]).to_pandas()
population_train_pd = star_schema.train.population.drop(["position"]).to_pandas() population_test_pd = star_schema.test.population.drop(["position"]).to_pandas()

In [30]:

  Copied!     
 
driver_standings_pd = driver_standings.drop(driver_standings.roles.unused).to_pandas()
lap_times_pd = lap_times.drop(lap_times.roles.unused).to_pandas()
pit_stops_pd = pit_stops.drop(pit_stops.roles.unused).to_pandas()
qualifying_pd = qualifying.drop(qualifying.roles.unused).to_pandas()
drivers_pd = drivers.drop(drivers.roles.unused).to_pandas()
driver_standings_pd = driver_standings.drop(driver_standings.roles.unused).to_pandas() lap_times_pd = lap_times.drop(lap_times.roles.unused).to_pandas() pit_stops_pd = pit_stops.drop(pit_stops.roles.unused).to_pandas() qualifying_pd = qualifying.drop(qualifying.roles.unused).to_pandas() drivers_pd = drivers.drop(drivers.roles.unused).to_pandas()

Because qualifying and drivers are many-to-one joins, we can directly join them onto our table.

In [31]:

  Copied!     
 
population_train_pd["id"] = population_train_pd.index

population_train_pd = population_train_pd.merge(
    qualifying_pd,
    on=["driverId", "raceId"],
    how="left",
)

population_train_pd = population_train_pd.merge(
    drivers_pd,
    on=["driverId"],
    how="left",
)

population_train_pd
population_train_pd["id"] = population_train_pd.index population_train_pd = population_train_pd.merge( qualifying_pd, on=["driverId", "raceId"], how="left", ) population_train_pd = population_train_pd.merge( drivers_pd, on=["driverId"], how="left", ) population_train_pd

Out[31]:

	raceId	driverId	year	win	date	id	qualifyId	position	number	nationality	driverRef
0	18	1	2008	1.0	2008-03-16	0	1	1.0	22.0	British	hamilton
1	18	2	2008	0.0	2008-03-16	1	5	5.0	3.0	German	heidfeld
2	18	3	2008	0.0	2008-03-16	2	7	7.0	7.0	German	rosberg
3	18	5	2008	0.0	2008-03-16	3	3	3.0	23.0	Finnish	kovalainen
4	18	6	2008	0.0	2008-03-16	4	14	14.0	8.0	Japanese	nakajima
...	...	...	...	...	...	...	...	...	...	...	...
25344	982	835	2017	0.0	2017-09-17	25344	7410	11.0	30.0	British	jolyon_palmer
25345	982	154	2017	0.0	2017-09-17	25345	7414	15.0	8.0	French	grosjean
25346	982	836	2017	0.0	2017-09-17	25346	7418	19.0	94.0	German	wehrlein
25347	982	18	2017	0.0	2017-09-17	25347	NaN	NaN	NaN	British	button
25348	982	814	2017	0.0	2017-09-17	25348	NaN	NaN	NaN	British	resta

25349 rows × 11 columns

Same for the testing set.

In [32]:

  Copied!     
 
population_test_pd["id"] = population_test_pd.index

population_test_pd = population_test_pd.merge(
    qualifying_pd,
    on=["driverId", "raceId"],
    how="left",
)

population_test_pd = population_test_pd.merge(
    drivers_pd,
    on=["driverId"],
    how="left",
)

population_test_pd
population_test_pd["id"] = population_test_pd.index population_test_pd = population_test_pd.merge( qualifying_pd, on=["driverId", "raceId"], how="left", ) population_test_pd = population_test_pd.merge( drivers_pd, on=["driverId"], how="left", ) population_test_pd

Out[32]:

	raceId	driverId	year	win	date	id	qualifyId	position	number	nationality	driverRef
0	18	4	2008	0.0	2008-03-16	0	12	12.0	5.0	Spanish	alonso
1	18	7	2008	0.0	2008-03-16	1	18	18.0	14.0	French	bourdais
2	19	3	2008	0.0	2008-03-23	2	38	16.0	7.0	German	rosberg
3	19	11	2008	0.0	2008-03-23	3	42	20.0	18.0	Japanese	sato
4	20	7	2008	0.0	2008-04-06	4	59	15.0	14.0	French	bourdais
...	...	...	...	...	...	...	...	...	...	...	...
6224	981	835	2017	0.0	2017-09-03	6224	7396	17.0	30.0	British	jolyon_palmer
6225	981	814	2017	0.0	2017-09-03	6225	NaN	NaN	NaN	British	resta
6226	982	822	2017	0.0	2017-09-17	6226	7405	6.0	77.0	Finnish	bottas
6227	982	832	2017	0.0	2017-09-17	6227	7409	10.0	55.0	Spanish	sainz
6228	982	826	2017	0.0	2017-09-17	6228	7412	13.0	26.0	Russian	kvyat

6229 rows × 11 columns

featuretools requires us to expand our peripheral tables so they can be joined using the unique id from the population table. Luckily, we can write a simple helper function that works for all three remaining peripheral tables.

In [33]:

  Copied!     
 
def prepare_peripheral(peripheral, train_or_test):
    """
    Helper function that imitates the behavior of 
    the data model defined above.
    """
    peripheral_new = peripheral.merge(
        train_or_test[["id", "driverId", "date"]],
        on=["driverId"],
    )

    peripheral_new = peripheral_new[
        peripheral_new["date_x"] < peripheral_new["date_y"]
    ]
    
    del peripheral_new["date_y"]
    del peripheral_new["driverId"]
    del peripheral_new["raceId"]

    return peripheral_new.rename(columns={"date_x": "date"})
def prepare_peripheral(peripheral, train_or_test): """ Helper function that imitates the behavior of the data model defined above. """ peripheral_new = peripheral.merge( train_or_test[["id", "driverId", "date"]], on=["driverId"], ) peripheral_new = peripheral_new[ peripheral_new["date_x"] < peripheral_new["date_y"] ] del peripheral_new["date_y"] del peripheral_new["driverId"] del peripheral_new["raceId"] return peripheral_new.rename(columns={"date_x": "date"})

In [34]:

  Copied!     
 
driver_standings_train_pd = prepare_peripheral(driver_standings_pd, population_train_pd)
driver_standings_test_pd = prepare_peripheral(driver_standings_pd, population_test_pd)
driver_standings_train_pd
driver_standings_train_pd = prepare_peripheral(driver_standings_pd, population_train_pd) driver_standings_test_pd = prepare_peripheral(driver_standings_pd, population_test_pd) driver_standings_train_pd

Out[34]:

	year	position	win	date	id
1	2008	1.0	1.0	2008-03-16	6
2	2008	1.0	1.0	2008-03-16	22
3	2008	1.0	1.0	2008-03-16	39
4	2008	1.0	1.0	2008-03-16	54
5	2008	1.0	1.0	2008-03-16	72
...	...	...	...	...	...
2653710	2017	19.0	0.0	2017-09-03	25344
2653802	2017	13.0	0.0	2017-09-03	25345
2653830	2017	16.0	0.0	2017-09-03	25346
2654096	2017	22.0	0.0	2017-09-03	25347
2654136	2017	23.0	0.0	2017-09-03	25348

1317941 rows × 5 columns

In [35]:

  Copied!     
 
lap_times_train_pd = prepare_peripheral(lap_times_pd, population_train_pd)
lap_times_test_pd = prepare_peripheral(lap_times_pd, population_test_pd)
lap_times_train_pd
lap_times_train_pd = prepare_peripheral(lap_times_pd, population_train_pd) lap_times_test_pd = prepare_peripheral(lap_times_pd, population_test_pd) lap_times_train_pd

Out[35]:

	year	lap	milliseconds	position	date	id
28	2009	1.0	109088.0	13.0	2009-03-29	5874
29	2009	1.0	109088.0	13.0	2009-03-29	5904
30	2009	1.0	109088.0	13.0	2009-03-29	5918
31	2009	1.0	109088.0	13.0	2009-03-29	5936
32	2009	1.0	109088.0	13.0	2009-03-29	5954
...	...	...	...	...	...	...
54748152	2017	49.0	85803.0	7.0	2017-09-03	25341
54748164	2017	50.0	85788.0	7.0	2017-09-03	25341
54748176	2017	51.0	85625.0	7.0	2017-09-03	25341
54748188	2017	52.0	86258.0	7.0	2017-09-03	25341
54748200	2017	53.0	86965.0	7.0	2017-09-03	25341

25544456 rows × 6 columns

lap_times demonstrates one of the greatest dangers of featuretools. Because it is written in pure Python, featuretools requires you expand your tables so that they can be more easily joined. But this comes at the cost of increased memory consumption: In this case a table that used to have about 420,000 rows has now been expanded to over 25 million rows.

In [36]:

  Copied!     
 
pit_stops_train_pd = prepare_peripheral(pit_stops_pd, population_train_pd)
pit_stops_test_pd = prepare_peripheral(pit_stops_pd, population_test_pd)
pit_stops_train_pd
pit_stops_train_pd = prepare_peripheral(pit_stops_pd, population_train_pd) pit_stops_test_pd = prepare_peripheral(pit_stops_pd, population_test_pd) pit_stops_train_pd

Out[36]:

	year	lap	milliseconds	stop	date	id
57	2011	16.0	23227.0	1.0	2011-03-27	22459
58	2011	16.0	23227.0	1.0	2011-03-27	22498
59	2011	16.0	23227.0	1.0	2011-03-27	22536
60	2011	16.0	23227.0	1.0	2011-03-27	22555
61	2011	16.0	23227.0	1.0	2011-03-27	22574
...	...	...	...	...	...	...
681497	2017	27.0	24173.0	2.0	2017-09-03	25332
681565	2017	25.0	31008.0	1.0	2017-09-03	25344
681593	2017	20.0	25212.0	1.0	2017-09-03	25346
681612	2017	16.0	24432.0	1.0	2017-09-03	25335
681624	2017	17.0	26131.0	1.0	2017-09-03	25341

213549 rows × 6 columns

In [37]:

  Copied!     
 
del population_train_pd["driverId"]
del population_train_pd["raceId"]
del population_train_pd["year"]
del population_train_pd["qualifyId"]
del population_train_pd["driverId"] del population_train_pd["raceId"] del population_train_pd["year"] del population_train_pd["qualifyId"]

In [38]:

  Copied!     
 
del population_test_pd["driverId"]
del population_test_pd["raceId"]
del population_test_pd["year"]
del population_test_pd["qualifyId"]
del population_test_pd["driverId"] del population_test_pd["raceId"] del population_test_pd["year"] del population_test_pd["qualifyId"]

In [39]:

  Copied!     
 
def add_index(df):
    df.insert(0, "index", range(len(df)))

population_pd_logical_types = {
    "id": ww.logical_types.Integer,
    "win": ww.logical_types.Integer,
    "date": ww.logical_types.Datetime,
    "position": ww.logical_types.IntegerNullable,
    "number": ww.logical_types.IntegerNullable,
    "nationality": ww.logical_types.Categorical,
    "driverRef": ww.logical_types.Categorical,
}
population_train_pd.ww.init(logical_types=population_pd_logical_types, index="id", name="population")
population_test_pd.ww.init(logical_types=population_pd_logical_types, index="id", name="population")

add_index(driver_standings_train_pd)
add_index(driver_standings_test_pd)
driver_standings_pd_logical_types = {
    "index": ww.logical_types.Integer,
    "year": ww.logical_types.Integer,
    "position": ww.logical_types.IntegerNullable,
    "win": ww.logical_types.Integer,
    "date": ww.logical_types.Datetime,
    "id": ww.logical_types.Integer,
}
driver_standings_train_pd.ww.init(logical_types=driver_standings_pd_logical_types, index="index", name="driver_standings")
driver_standings_test_pd.ww.init(logical_types=driver_standings_pd_logical_types, index="index", name="driver_standings")

add_index(lap_times_train_pd)
add_index(lap_times_test_pd)
lap_times_pd_logical_types = {
    "index": ww.logical_types.Integer,
    "year": ww.logical_types.Integer,
    "lap": ww.logical_types.Integer,
    "milliseconds": ww.logical_types.Integer,
    "position": ww.logical_types.IntegerNullable,
    "date": ww.logical_types.Datetime,
    "id": ww.logical_types.Integer,
}
lap_times_train_pd.ww.init(logical_types=lap_times_pd_logical_types, index="index", name="lap_times")
lap_times_test_pd.ww.init(logical_types=lap_times_pd_logical_types, index="index", name="lap_times")

add_index(pit_stops_train_pd)
add_index(pit_stops_test_pd)
pit_stops_pd_logical_types = {
    "index": ww.logical_types.Integer,
    "year": ww.logical_types.Integer,
    "lap": ww.logical_types.Integer,
    "milliseconds": ww.logical_types.Integer,
    "stop": ww.logical_types.Categorical,
    "date": ww.logical_types.Datetime,
    "id": ww.logical_types.Integer,
}
pit_stops_train_pd.ww.init(logical_types=pit_stops_pd_logical_types, index="index", name="pit_stops")
pit_stops_test_pd.ww.init(logical_types=pit_stops_pd_logical_types, index="index", name="pit_stops")
def add_index(df): df.insert(0, "index", range(len(df))) population_pd_logical_types = { "id": ww.logical_types.Integer, "win": ww.logical_types.Integer, "date": ww.logical_types.Datetime, "position": ww.logical_types.IntegerNullable, "number": ww.logical_types.IntegerNullable, "nationality": ww.logical_types.Categorical, "driverRef": ww.logical_types.Categorical, } population_train_pd.ww.init(logical_types=population_pd_logical_types, index="id", name="population") population_test_pd.ww.init(logical_types=population_pd_logical_types, index="id", name="population") add_index(driver_standings_train_pd) add_index(driver_standings_test_pd) driver_standings_pd_logical_types = { "index": ww.logical_types.Integer, "year": ww.logical_types.Integer, "position": ww.logical_types.IntegerNullable, "win": ww.logical_types.Integer, "date": ww.logical_types.Datetime, "id": ww.logical_types.Integer, } driver_standings_train_pd.ww.init(logical_types=driver_standings_pd_logical_types, index="index", name="driver_standings") driver_standings_test_pd.ww.init(logical_types=driver_standings_pd_logical_types, index="index", name="driver_standings") add_index(lap_times_train_pd) add_index(lap_times_test_pd) lap_times_pd_logical_types = { "index": ww.logical_types.Integer, "year": ww.logical_types.Integer, "lap": ww.logical_types.Integer, "milliseconds": ww.logical_types.Integer, "position": ww.logical_types.IntegerNullable, "date": ww.logical_types.Datetime, "id": ww.logical_types.Integer, } lap_times_train_pd.ww.init(logical_types=lap_times_pd_logical_types, index="index", name="lap_times") lap_times_test_pd.ww.init(logical_types=lap_times_pd_logical_types, index="index", name="lap_times") add_index(pit_stops_train_pd) add_index(pit_stops_test_pd) pit_stops_pd_logical_types = { "index": ww.logical_types.Integer, "year": ww.logical_types.Integer, "lap": ww.logical_types.Integer, "milliseconds": ww.logical_types.Integer, "stop": ww.logical_types.Categorical, "date": ww.logical_types.Datetime, "id": ww.logical_types.Integer, } pit_stops_train_pd.ww.init(logical_types=pit_stops_pd_logical_types, index="index", name="pit_stops") pit_stops_test_pd.ww.init(logical_types=pit_stops_pd_logical_types, index="index", name="pit_stops")

In [40]:

  Copied!     
 
dataframes_train = {
    "population" : (population_train_pd, ),
    "driver_standings": (driver_standings_train_pd, ),
    "lap_times" : (lap_times_train_pd, ),
    "pit_stops" : (pit_stops_train_pd, ),
}
dataframes_train = { "population" : (population_train_pd, ), "driver_standings": (driver_standings_train_pd, ), "lap_times" : (lap_times_train_pd, ), "pit_stops" : (pit_stops_train_pd, ), }

In [41]:

  Copied!     
 
dataframes_test = {
    "population" : (population_test_pd, ),
    "driver_standings": (driver_standings_test_pd, ),
    "lap_times" : (lap_times_test_pd, ),
    "pit_stops" : (pit_stops_test_pd, )
}
dataframes_test = { "population" : (population_test_pd, ), "driver_standings": (driver_standings_test_pd, ), "lap_times" : (lap_times_test_pd, ), "pit_stops" : (pit_stops_test_pd, ) }

In [42]:

  Copied!     
 
relationships = [
    ("population", "id", "driver_standings", "id"),
    ("population", "id", "lap_times", "id"),
    ("population", "id", "pit_stops", "id")
]
relationships = [ ("population", "id", "driver_standings", "id"), ("population", "id", "lap_times", "id"), ("population", "id", "pit_stops", "id") ]

In [43]:

  Copied!     
 
featuretools_train_pd = featuretools.dfs(
    dataframes=dataframes_train,
    relationships=relationships,
    target_dataframe_name="population")[0]
featuretools_train_pd = featuretools.dfs( dataframes=dataframes_train, relationships=relationships, target_dataframe_name="population")[0]

In [44]:

  Copied!     
 
featuretools_test_pd = featuretools.dfs(
    dataframes=dataframes_test,
    relationships=relationships,
    target_dataframe_name="population")[0]
featuretools_test_pd = featuretools.dfs( dataframes=dataframes_test, relationships=relationships, target_dataframe_name="population")[0]

In [45]:

  Copied!     
 
featuretools_train = getml.data.DataFrame.from_pandas(featuretools_train_pd, "featuretools_train")
featuretools_test = getml.data.DataFrame.from_pandas(featuretools_test_pd, "featuretools_test")
featuretools_train = getml.data.DataFrame.from_pandas(featuretools_train_pd, "featuretools_train") featuretools_test = getml.data.DataFrame.from_pandas(featuretools_test_pd, "featuretools_test")

In [46]:

  Copied!     
 
featuretools_train.set_role("win", getml.data.roles.target)
featuretools_train.set_role(featuretools_train.roles.unused_float, getml.data.roles.numerical)
featuretools_train.set_role(featuretools_train.roles.unused_string, getml.data.roles.categorical)

featuretools_train
featuretools_train.set_role("win", getml.data.roles.target) featuretools_train.set_role(featuretools_train.roles.unused_float, getml.data.roles.numerical) featuretools_train.set_role(featuretools_train.roles.unused_string, getml.data.roles.categorical) featuretools_train

Out[46]:

name	win	position	number	nationality	driverRef	COUNT(driver_standings)	COUNT(lap_times)	COUNT(pit_stops)	MODE(pit_stops.stop)	NUM_UNIQUE(pit_stops.stop)	DAY(date)	MONTH(date)	WEEKDAY(date)	YEAR(date)	MODE(driver_standings.DAY(date))	MODE(driver_standings.MONTH(date))	MODE(driver_standings.WEEKDAY(date))	MODE(driver_standings.YEAR(date))	NUM_UNIQUE(driver_standings.DAY(date))	NUM_UNIQUE(driver_standings.MONTH(date))	NUM_UNIQUE(driver_standings.WEEKDAY(date))	NUM_UNIQUE(driver_standings.YEAR(date))	MODE(lap_times.DAY(date))	MODE(lap_times.MONTH(date))	MODE(lap_times.WEEKDAY(date))	MODE(lap_times.YEAR(date))	NUM_UNIQUE(lap_times.DAY(date))	NUM_UNIQUE(lap_times.MONTH(date))	NUM_UNIQUE(lap_times.WEEKDAY(date))	NUM_UNIQUE(lap_times.YEAR(date))	MODE(pit_stops.DAY(date))	MODE(pit_stops.MONTH(date))	MODE(pit_stops.WEEKDAY(date))	MODE(pit_stops.YEAR(date))	NUM_UNIQUE(pit_stops.DAY(date))	NUM_UNIQUE(pit_stops.MONTH(date))	NUM_UNIQUE(pit_stops.WEEKDAY(date))	NUM_UNIQUE(pit_stops.YEAR(date))	MAX(driver_standings.position)	MAX(driver_standings.win)	MAX(driver_standings.year)	MEAN(driver_standings.position)	MEAN(driver_standings.win)	MEAN(driver_standings.year)	MIN(driver_standings.position)	MIN(driver_standings.win)	MIN(driver_standings.year)	SKEW(driver_standings.position)	SKEW(driver_standings.win)	SKEW(driver_standings.year)	STD(driver_standings.position)	STD(driver_standings.win)	STD(driver_standings.year)	SUM(driver_standings.position)	SUM(driver_standings.win)	SUM(driver_standings.year)	MAX(lap_times.lap)	MAX(lap_times.milliseconds)	MAX(lap_times.position)	MAX(lap_times.year)	MEAN(lap_times.lap)	MEAN(lap_times.milliseconds)	MEAN(lap_times.position)	MEAN(lap_times.year)	MIN(lap_times.lap)	MIN(lap_times.milliseconds)	MIN(lap_times.position)	MIN(lap_times.year)	SKEW(lap_times.lap)	SKEW(lap_times.milliseconds)	SKEW(lap_times.position)	SKEW(lap_times.year)	STD(lap_times.lap)	STD(lap_times.milliseconds)	STD(lap_times.position)	STD(lap_times.year)	SUM(lap_times.lap)	SUM(lap_times.milliseconds)	SUM(lap_times.position)	SUM(lap_times.year)	MAX(pit_stops.lap)	MAX(pit_stops.milliseconds)	MAX(pit_stops.year)	MEAN(pit_stops.lap)	MEAN(pit_stops.milliseconds)	MEAN(pit_stops.year)	MIN(pit_stops.lap)	MIN(pit_stops.milliseconds)	MIN(pit_stops.year)	SKEW(pit_stops.lap)	SKEW(pit_stops.milliseconds)	SKEW(pit_stops.year)	STD(pit_stops.lap)	STD(pit_stops.milliseconds)	STD(pit_stops.year)	SUM(pit_stops.lap)	SUM(pit_stops.milliseconds)	SUM(pit_stops.year)
role	target	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical
0	1	1	22	British	hamilton	17	1037	0	NULL	NULL	16	3	6	2008	8	7	6	2007	16	8	1	1	8	7	6	2007	16	8	1	1	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	3	1	2007	1.4706	0.2353	2007	1	0	2007	1.3538	1.3723	0	0.7998	0.4372	0	25	4	34119	78	1453884	19	2007	32.0559	90600.8602	3.2575	2007	1	72506	1	2007	0.1882	26.093	2.4493	0	19.1156	45564.9471	3.4488	0	33242	93953092	3378	2081259	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0
1	0	5	3	German	heidfeld	134	6863	0	NULL	NULL	16	3	6	2008	1	7	6	2005	31	8	1	8	1	7	6	2007	31	8	1	8	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	21	0	2007	10.8582	0	2003.5448	4	0	2000	0.4681	0	-0.04499	5.136	0	2.3255	1455	0	268475	78	1285986	22	2007	30.4682	89072.7102	9.6799	2003.7322	1	70516	1	2000	0.2623	30.5344	0.3339	-0.05998	18.7041	20544.1746	4.199	2.258	209103	611306010	66433	13751614	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0
2	0	7	7	German	rosberg	35	1627	0	NULL	NULL	16	3	6	2008	8	7	6	2006	25	8	1	2	8	5	6	2007	25	8	1	2	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	17	0	2007	11.9714	0	2006.4857	7	0	2006	0.1328	0	0.05976	3.1761	0	0.5071	419	0	70227	77	191891	22	2007	30.4124	89820.1174	9.9809	2006.5931	1	73159	3	2006	0.2225	1.7858	0.5491	-0.3794	18.6109	14212.8848	3.733	0.4914	49481	146137331	16239	3264727	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0
3	0	3	23	Finnish	kovalainen	17	1023	0	NULL	NULL	16	3	6	2008	8	7	6	2007	16	8	1	1	8	7	6	2007	16	8	1	1	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	11	0	2007	8.3529	0	2007	7	0	2007	0.7931	0	0	1.4116	0	0	142	0	34119	76	1394884	22	2007	31.4526	92458.8299	9.4018	2007	1	73998	1	2007	0.1906	26.2493	0.6941	0	18.6943	43578.4946	4.1724	0	32176	94585383	9618	2053161	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0
4	0	14	8	Japanese	nakajima	1	70	0	NULL	NULL	16	3	6	2008	21	10	6	2007	1	1	1	1	21	10	6	2007	1	1	1	1	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	22	0	2007	22	0	2007	22	0	2007	nan	nan	nan	nan	nan	nan	22	0	2007	70	104396	16	2007	35.5	76228.3571	10.9143	2007	1	73116	9	2007	0	4.4222	1.2783	0	20.3511	4785.2529	1.8552	0	2485	5335985	764	140490	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0
	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
25344	0	11	30	British	jolyon_palmer	34	1517	56	1	4	17	9	6	2017	3	7	6	2016	25	9	1	2	9	7	6	2016	25	9	1	2	9	7	6	2016	21	9	1	2	20	0	2017	18.0588	0	2016.3824	11	0	2016	-1.7626	0	0.5068	1.8413	0	0.4933	614	0	68557	78	1223356	22	2017	29.4199	98411.6038	14.5175	2016.3665	1	68652	7	2016	0.2921	21.0237	0.3751	0.5546	18.5854	43393.1519	3.0483	0.482	44630	149290403	22023	3058828	50	1088294	2017	20.9286	61015.1429	2016.2857	1	16851	2016	0.3035	5.1507	0.975	13.473	191214.0556	0.4558	1172	3416848	112912
25345	0	15	8	French	grosjean	118	5518	206	1	5	17	9	6	2017	27	7	6	2016	31	9	1	7	9	7	6	2016	31	9	1	7	9	7	6	2014	31	9	1	6	24	0	2017	11.5254	0	2014.0508	5	0	2009	1.3306	0	-0.5931	4.1709	0	2.0542	1360	0	237658	78	3670560	24	2017	29.4511	99489.4235	10.7356	2014.1414	1	68590	1	2009	0.2565	39.9299	0.06907	-0.6145	18.0521	63108.3058	4.8371	1.9904	162511	548982639	59239	11114032	62	1313665	2017	22.9369	46260.6214	2014.4272	1	15073	2012	0.3075	7.1215	0.04707	13.7483	154388.9496	1.6024	4725	9529688	414972
25346	0	19	94	German	wehrlein	32	1636	62	1	5	17	9	6	2017	3	7	6	2016	24	9	1	2	3	7	6	2016	24	9	1	2	3	7	6	2016	24	8	1	2	22	0	2017	17.25	0	2016.3438	13	0	2016	0.3213	0	0.6908	2.2718	0	0.4826	552	0	64523	76	2118880	22	2017	30.3839	99952.8594	15.8319	2016.3594	1	69241	4	2016	0.2027	20.0905	-0.8509	0.5865	18.2369	77850.4087	3.2136	0.48	49708	163522878	25901	3298764	52	2008464	2017	21.6129	119925.5968	2016.3065	1	18064	2016	0.1765	3.9986	0.8606	12.8896	368710.9372	0.4648	1340	7435387	125011
25347	0	NULL	NULL	British	button	309	16272	259	1	6	17	9	6	2017	24	7	6	2016	31	9	1	18	24	7	6	2013	31	9	1	18	25	7	6	2011	31	9	1	7	22	1	2017	9.6828	0.04854	2008.6181	1	0	2000	0.3055	4.2218	-0.099	5.8899	0.2153	5.0436	2992	15	620663	78	7503775	22	2017	30.2853	95662.1676	8.4123	2008.4672	1	69828	1	2000	0.2244	64.7495	0.4521	-0.1278	18.3544	79242.6165	4.859	4.8663	492803	1556614792	136885	32681778	64	2010062	2017	24.9344	45562.5985	2013.3977	1	14501	2011	0.3229	8.7373	0.09889	13.5509	179696.8608	1.785	6458	11800713	521470
25348	0	NULL	NULL	British	resta	61	3284	127	1	4	17	9	6	2017	27	7	6	2012	27	9	1	4	27	7	6	2012	27	9	1	4	25	5	6	2011	27	9	1	4	23	0	2017	12.6066	0	2012.2459	8	0	2011	1.244	0	2.0563	3.4558	0	1.3498	769	0	122747	78	7503980	24	2017	30.3097	103049.4318	10.7728	2012.0539	1	76855	1	2011	0.1989	44.5439	0.1348	1.8215	18.0742	144158.4761	3.5107	1.0439	99537	338414334	35378	6607585	57	123124	2017	25.1496	24084.2598	2011.9528	1	14538	2011	0.3166	8.1337	1.2541	13.8456	9973.774	0.9416	3194	3058701	255518

25349 rows x 98 columns
memory usage: 16.12 MB
name: featuretools_train
type: getml.DataFrame

In [47]:

  Copied!     
 
featuretools_test.set_role("win", getml.data.roles.target)
featuretools_test.set_role(featuretools_test.roles.unused_float, getml.data.roles.numerical)
featuretools_test.set_role(featuretools_test.roles.unused_string, getml.data.roles.categorical)

featuretools_test
featuretools_test.set_role("win", getml.data.roles.target) featuretools_test.set_role(featuretools_test.roles.unused_float, getml.data.roles.numerical) featuretools_test.set_role(featuretools_test.roles.unused_string, getml.data.roles.categorical) featuretools_test

Out[47]:

name	win	position	number	nationality	driverRef	COUNT(driver_standings)	COUNT(lap_times)	COUNT(pit_stops)	MODE(pit_stops.stop)	NUM_UNIQUE(pit_stops.stop)	DAY(date)	MONTH(date)	WEEKDAY(date)	YEAR(date)	MODE(driver_standings.DAY(date))	MODE(driver_standings.MONTH(date))	MODE(driver_standings.WEEKDAY(date))	MODE(driver_standings.YEAR(date))	NUM_UNIQUE(driver_standings.DAY(date))	NUM_UNIQUE(driver_standings.MONTH(date))	NUM_UNIQUE(driver_standings.WEEKDAY(date))	NUM_UNIQUE(driver_standings.YEAR(date))	MODE(lap_times.DAY(date))	MODE(lap_times.MONTH(date))	MODE(lap_times.WEEKDAY(date))	MODE(lap_times.YEAR(date))	NUM_UNIQUE(lap_times.DAY(date))	NUM_UNIQUE(lap_times.MONTH(date))	NUM_UNIQUE(lap_times.WEEKDAY(date))	NUM_UNIQUE(lap_times.YEAR(date))	MODE(pit_stops.DAY(date))	MODE(pit_stops.MONTH(date))	MODE(pit_stops.WEEKDAY(date))	MODE(pit_stops.YEAR(date))	NUM_UNIQUE(pit_stops.DAY(date))	NUM_UNIQUE(pit_stops.MONTH(date))	NUM_UNIQUE(pit_stops.WEEKDAY(date))	NUM_UNIQUE(pit_stops.YEAR(date))	MAX(driver_standings.position)	MAX(driver_standings.win)	MAX(driver_standings.year)	MEAN(driver_standings.position)	MEAN(driver_standings.win)	MEAN(driver_standings.year)	MIN(driver_standings.position)	MIN(driver_standings.win)	MIN(driver_standings.year)	SKEW(driver_standings.position)	SKEW(driver_standings.win)	SKEW(driver_standings.year)	STD(driver_standings.position)	STD(driver_standings.win)	STD(driver_standings.year)	SUM(driver_standings.position)	SUM(driver_standings.win)	SUM(driver_standings.year)	MAX(lap_times.lap)	MAX(lap_times.milliseconds)	MAX(lap_times.position)	MAX(lap_times.year)	MEAN(lap_times.lap)	MEAN(lap_times.milliseconds)	MEAN(lap_times.position)	MEAN(lap_times.year)	MIN(lap_times.lap)	MIN(lap_times.milliseconds)	MIN(lap_times.position)	MIN(lap_times.year)	SKEW(lap_times.lap)	SKEW(lap_times.milliseconds)	SKEW(lap_times.position)	SKEW(lap_times.year)	STD(lap_times.lap)	STD(lap_times.milliseconds)	STD(lap_times.position)	STD(lap_times.year)	SUM(lap_times.lap)	SUM(lap_times.milliseconds)	SUM(lap_times.position)	SUM(lap_times.year)	MAX(pit_stops.lap)	MAX(pit_stops.milliseconds)	MAX(pit_stops.year)	MEAN(pit_stops.lap)	MEAN(pit_stops.milliseconds)	MEAN(pit_stops.year)	MIN(pit_stops.lap)	MIN(pit_stops.milliseconds)	MIN(pit_stops.year)	SKEW(pit_stops.lap)	SKEW(pit_stops.milliseconds)	SKEW(pit_stops.year)	STD(pit_stops.lap)	STD(pit_stops.milliseconds)	STD(pit_stops.year)	SUM(pit_stops.lap)	SUM(pit_stops.milliseconds)	SUM(pit_stops.year)
role	target	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	categorical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical	numerical
0	0	12	5	Spanish	alonso	105	5769	0	NULL	NULL	16	3	6	2008	1	7	6	2005	31	8	1	6	24	7	6	2006	31	8	1	6	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	23	1	2007	5.6667	0.181	2004.3714	1	0	2001	1.7319	1.6816	-0.3952	7.1598	0.3868	1.9575	595	19	210459	78	1406432	21	2007	30.436	88439.8289	5.2479	2004.5883	1	70526	1	2001	0.2301	34.9402	1.3932	-0.4971	18.3774	22633.3599	4.9145	1.8675	175585	510209373	30275	11564470	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0
1	0	18	14	French	bourdais	0	0	0	NULL	NULL	16	3	6	2008	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0	0	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0
2	0	16	7	German	rosberg	36	1685	0	NULL	NULL	23	3	6	2008	8	7	6	2006	25	8	1	3	16	5	6	2007	25	8	1	3	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	17	0	2008	11.7222	0	2006.5278	3	0	2006	-0.1958	0	0.4024	3.4691	0	0.5599	422	0	72235	77	191891	22	2008	30.381	90110.451	9.7822	2006.6415	1	73159	2	2006	0.2179	1.8323	0.5148	0.05473	18.5504	14476.8352	3.8237	0.5468	51192	151836110	16483	3381191	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0
3	0	20	18	Japanese	sato	84	4309	0	NULL	NULL	23	3	6	2008	12	7	6	2004	31	8	1	6	25	7	6	2007	31	8	1	7	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	23	0	2007	16.9405	0	2004.8095	7	0	2002	-0.493	0	-0.3829	5.5479	0	1.7319	1423	0	168404	76	1921214	22	2008	29.1091	91287.2075	12.4342	2004.8867	1	70727	1	2002	0.2832	32.4719	-0.2034	-0.3841	18.0839	42719.4804	4.861	1.7193	125431	393356577	53579	8639057	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0
4	0	15	14	French	bourdais	2	55	0	NULL	NULL	6	4	6	2008	16	3	6	2008	2	1	1	1	16	3	6	2008	1	1	1	1	NULL	NULL	NULL	NULL	NULL	NULL	NULL	NULL	10	0	2008	8.5	0	2008	7	0	2008	nan	nan	nan	2.1213	0	0	17	0	4016	55	156156	16	2008	28	98864.3818	9.7818	2008	1	89534	4	2008	0	2.2604	-0.06348	0	16.0208	17491.8075	4.3149	0	1540	5437541	538	110440	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	nan	0	0	0
	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
6224	0	17	30	British	jolyon_palmer	33	1488	55	1	4	3	9	6	2017	9	7	6	2016	25	9	1	2	9	7	6	2016	25	9	1	2	9	7	6	2016	21	9	1	2	20	0	2017	18.0303	0	2016.3636	11	0	2016	-1.7159	0	0.5943	1.8622	0	0.4885	595	0	66540	78	1223356	22	2017	29.7009	98587.2056	14.5195	2016.3542	1	68652	7	2016	0.2691	20.8455	0.3776	0.6105	18.6186	43789.6099	3.0523	0.4784	44195	146697762	21605	3000335	50	1088294	2017	20.8545	61560.7273	2016.2727	1	16851	2016	0.3181	5.1018	1.0495	13.5857	192932.4401	0.4495	1147	3385840	110895
6225	0	NULL	NULL	British	resta	60	3284	127	1	4	3	9	6	2017	27	7	6	2012	27	9	1	4	27	7	6	2012	27	9	1	4	25	5	6	2011	27	9	1	4	23	0	2017	12.4333	0	2012.1667	8	0	2011	1.155	0	2.0461	3.2068	0	1.2097	746	0	120730	78	7503980	24	2017	30.3097	103049.4318	10.7728	2012.0539	1	76855	1	2011	0.1989	44.5439	0.1348	1.8215	18.0742	144158.4761	3.5107	1.0439	99537	338414334	35378	6607585	57	123124	2017	25.1496	24084.2598	2011.9528	1	14538	2011	0.3166	8.1337	1.2541	13.8456	9973.774	0.9416	3194	3058701	255518
6226	0	6	77	Finnish	bottas	91	5153	183	1	5	17	9	6	2017	9	7	6	2016	31	9	1	5	9	7	6	2016	31	9	1	5	9	7	6	2016	31	9	1	5	18	1	2017	8.1868	0.02198	2014.8901	3	0	2013	1.0124	6.6307	0.04113	4.9976	0.1474	1.3618	745	2	183355	78	3689718	22	2017	30.4306	99678.869	8.0584	2014.8894	1	67847	1	2013	0.2271	32.1149	0.5698	0.05111	18.18	74208.5105	4.885	1.3584	156809	513645212	41525	10382725	59	2008595	2017	24.1093	68069.5082	2014.8087	1	14978	2013	0.3284	5.9016	0.07557	13.4151	242520.4324	1.3349	4412	12456720	368710
6227	0	10	55	Spanish	sainz	53	2690	97	1	4	17	9	6	2017	3	7	6	2016	29	9	1	3	30	5	6	2016	29	9	1	3	9	7	6	2016	29	9	1	3	16	0	2017	11.8679	0	2015.8868	7	0	2015	0.06274	0	0.2015	2.481	0	0.7761	629	0	106842	78	2122068	21	2017	29.7457	99861.4349	10.655	2015.8587	1	69150	3	2015	0.3078	24.5072	0.3527	0.2326	18.3133	61998.6099	3.5636	0.7425	80016	268627260	28662	5422660	51	2008828	2017	22.6289	84961.6907	2015.8247	1	14966	2015	0.3388	5.1188	0.2777	12.7919	297744.5008	0.7218	2195	8241284	195535
6228	0	13	26	Russian	kvyat	72	3585	145	1	5	17	9	6	2017	9	7	6	2016	31	9	1	4	30	7	6	2015	31	9	1	4	9	7	6	2016	31	9	1	4	22	0	2017	12.6111	0	2015.3889	7	0	2014	-0.3133	0	0.08466	3.4214	0	1.0688	908	0	145108	78	3690660	22	2017	29.2739	100722.5623	11.2388	2015.3819	1	68061	2	2014	0.2624	33.405	-0.07759	0.126	18.0505	79720.0053	4.0792	1.0596	104947	361090386	40291	7225144	59	2011266	2017	23.6207	64171.0069	2015.3586	1	16139	2014	0.2769	6.5459	0.07908	15.0675	236423.1748	1.0651	3425	9304796	292227

6229 rows x 98 columns
memory usage: 3.96 MB
name: featuretools_test
type: getml.DataFrame

We train an untuned XGBoostRegressor on top of featuretools' features, just like we have done for getML's features.

Since some of featuretools features are categorical, we allow the pipeline to include these features as well. Other features contain NaN values, which is why we also apply getML's Imputation preprocessor.

In [48]:

  Copied!     
 
imputation = getml.preprocessors.Imputation()

predictor = getml.predictors.XGBoostClassifier(n_jobs=1)

pipe2 = getml.pipeline.Pipeline(
    tags=['featuretools'],
    preprocessors=[imputation],
    predictors=[predictor],
    include_categorical=True,
)

pipe2
imputation = getml.preprocessors.Imputation() predictor = getml.predictors.XGBoostClassifier(n_jobs=1) pipe2 = getml.pipeline.Pipeline( tags=['featuretools'], preprocessors=[imputation], predictors=[predictor], include_categorical=True, ) pipe2

Out[48]:

Pipeline(data_model='population',
         feature_learners=[],
         feature_selectors=[],
         include_categorical=True,
         loss_function='SquareLoss',
         peripheral=[],
         predictors=['XGBoostClassifier'],
         preprocessors=['Imputation'],
         share_selected_features=0.5,
         tags=['featuretools'])

In [49]:

  Copied!     
 
pipe2.fit(featuretools_train)
pipe2.fit(featuretools_train)

Checking data model...

  Staging... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00
  Preprocessing... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00
  Checking... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00

The pipeline check generated 1 issues labeled INFO and 16 issues labeled WARNING.

To see the issues in full, run .check() on the pipeline.

  Staging... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00
  Preprocessing... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00
  XGBoost: Training as predictor... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:14

Trained pipeline.

Time taken: 0:00:14.526013.

Out[49]:

Pipeline(data_model='population',
         feature_learners=[],
         feature_selectors=[],
         include_categorical=True,
         loss_function='SquareLoss',
         peripheral=[],
         predictors=['XGBoostClassifier'],
         preprocessors=['Imputation'],
         share_selected_features=0.5,
         tags=['featuretools'])

In [50]:

  Copied!     
 
featuretools_score = pipe2.score(featuretools_test)
featuretools_score
featuretools_score = pipe2.score(featuretools_test) featuretools_score

  Staging... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00
  Preprocessing... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% • 00:00

Out[50]:

	date time	set used	target	accuracy	auc	cross entropy
0	2024-09-12 12:54:50	featuretools_train	win	0.9717	0.9389	0.08364
1	2024-09-12 12:54:50	featuretools_test	win	0.9722	0.9153	0.08797

2.6 Studying features¶

We would like to understand why getML outperforms featuretools. In particular, getML's FastProp is based on an approach that is very similar to featuretools.

To investigate this matter, we first take a look at the importance of the features FastProp has learned:

In [51]:

  Copied!     
 
names, importances = pipe1.features.importances(target_num=0)

plt.subplots(figsize=(20, 10))

plt.bar(names[:30], importances[:30])

plt.title("feature importances")
plt.grid(True)
plt.xlabel("column")
plt.ylabel("importance")
plt.xticks(rotation='vertical')

plt.show()
names, importances = pipe1.features.importances(target_num=0) plt.subplots(figsize=(20, 10)) plt.bar(names[:30], importances[:30]) plt.title("feature importances") plt.grid(True) plt.xlabel("column") plt.ylabel("importance") plt.xticks(rotation='vertical') plt.show()

No description has been provided for this image

We take a look at the most important features, to get an idea where the predictive power comes from:

In [52]:

  Copied!     
 
pipe1.features.to_sql().find(names[0])[0]
pipe1.features.to_sql().find(names[0])[0]

Out[52]:

DROP TABLE IF EXISTS "T4__DRIVERREF__MAPPING_TARGET_1_AVG";

CREATE TABLE "T4__DRIVERREF__MAPPING_TARGET_1_AVG"("key" TEXT, "value" REAL);

INSERT INTO "T4__DRIVERREF__MAPPING_TARGET_1_AVG" ("key", "value")
VALUES('fangio', 0.3653846153846154),
      ('ascari', 0.3055555555555556),
      ('michael_schumacher', 0.3032786885245902),
      ('hamilton', 0.2981366459627329),
      ('stewart', 0.2948717948717949),
      ('prost', 0.264367816091954),
      ('vettel', 0.2467532467532468),
      ('clark', 0.2394366197183098),
      ('senna', 0.2323943661971831),
      ('damon_hill', 0.2159090909090909),
      ('moss', 0.1866666666666667),
      ('mansell', 0.1569767441860465),
      ('lauda', 0.1390728476821192),
      ('alonso', 0.1282051282051282),
      ('hakkinen', 0.1278195488721804),
      ('rosberg', 0.10625),
      ('hunt', 0.1058823529411765),
      ('jack_brabham', 0.1052631578947368),
      ('piquet', 0.09714285714285714),
      ('farina', 0.09523809523809523),
      ('jones', 0.09259259259259259),
      ('brooks', 0.09090909090909091),
      ('scheckter', 0.09),
      ('montoya', 0.08536585365853659),
      ('emerson_fittipaldi', 0.08403361344537816),
      ('rindt', 0.08333333333333333),
      ('hulme', 0.08247422680412371),
      ('peterson', 0.08247422680412371),
      ('reutemann', 0.07936507936507936),
      ('mario_andretti', 0.07142857142857142),
      ('raikkonen', 0.0684931506849315),
      ('villeneuve', 0.06153846153846154),
      ('surtees', 0.0576923076923077),
      ('hill', 0.05194805194805195),
      ('gilles_villeneuve', 0.05),
      ('berger', 0.05),
      ('keke_rosberg', 0.04901960784313725),
      ('hawthorn', 0.04878048780487805),
      ('ickx', 0.04878048780487805),
      ('gurney', 0.04819277108433735),
      ('collins', 0.04761904761904762),
      ('arnoux', 0.04697986577181208),
      ('massa', 0.04587155963302753),
      ('button', 0.04511278195488722),
      ('pironi', 0.04347826086956522),
      ('coulthard', 0.0427807486631016),
      ('revson', 0.0425531914893617),
      ('ricciardo', 0.0425531914893617),
      ('webber', 0.04216867469879518),
      ('ralf_schumacher', 0.03846153846153846),
      ('barrichello', 0.036),
      ('phil_hill', 0.03448275862068965),
      ('rodriguez', 0.03333333333333333),
      ('scarfiotti', 0.03333333333333333),
      ('laffite', 0.03311258278145696),
      ('regazzoni', 0.03225806451612903),
      ('jabouille', 0.03076923076923077),
      ('baghetti', 0.02941176470588235),
      ('flaherty', 0.02941176470588235),
      ('watson', 0.02898550724637681),
      ('mclaren', 0.02803738317757009),
      ('gethin', 0.02777777777777778),
      ('musso', 0.02702702702702703),
      ('cevert', 0.02631578947368421),
      ('ruttman', 0.025),
      ('irvine', 0.025),
      ('max_verstappen', 0.02380952380952381),
      ('hanks', 0.02325581395348837),
      ('alboreto', 0.02298850574712644),
      ('boutsen', 0.02290076335877863),
      ('herbert', 0.02222222222222222),
      ('parsons', 0.02083333333333333),
      ('bryan', 0.02040816326530612),
      ('trintignant', 0.02),
      ('kubica', 0.0196078431372549),
      ('patrese', 0.01951219512195122),
      ('rathmann', 0.01923076923076923),
      ('bandini', 0.01851851851851852),
      ('tambay', 0.01834862385321101),
      ('ginther', 0.01639344262295082),
      ('fisichella', 0.01621621621621622),
      ('ward', 0.01612903225806452),
      ('frentzen', 0.01574803149606299),
      ('brambilla', 0.01492537313432836),
      ('bottas', 0.01470588235294118),
      ('pace', 0.0136986301369863),
      ('maldonado', 0.01298701298701299),
      ('beltoise', 0.01282051282051282),
      ('siffert', 0.01265822784810127),
      ('depailler', 0.0108695652173913),
      ('angelis', 0.0108695652173913),
      ('bonnier', 0.009009009009009009),
      ('mass', 0.008771929824561403),
      ('panis', 0.007692307692307693),
      ('darter', 0),
      ('pretorius', 0),
      ('adamich', 0),
      ('trevor_taylor', 0),
      ('love', 0),
      ('anderson', 0),
      ('bianchi', 0),
      ('courage', 0),
      ('tingle', 0),
      ('attwood', 0),
      ('spence', 0),
      ('resta', 0),
      ('chiron', 0),
      ('reece', 0),
      ('linden', 0),
      ('agabashian', 0),
      ('manzon', 0),
      ('rosier', 0),
      ('villoresi', 0),
      ('graffenried', 0),
      ('hulkenberg', 0),
      ('petrov', 0),
      ('bruno_senna', 0),
      ('daywalt', 0),
      ('perez', 0),
      ('vergne', 0),
      ('pic', 0),
      ('chilton', 0),
      ('gutierrez', 0),
      ('jules_bianchi', 0),
      ('kevin_magnussen', 0),
      ('kvyat', 0),
      ('ericsson', 0),
      ('nasr', 0),
      ('sainz', 0),
      ('schell', 0),
      ('maggs', 0),
      ('gregory', 0),
      ('andre_pilette', 0),
      ('beaufort', 0),
      ('burgess', 0),
      ('salvadori', 0),
      ('trips', 0),
      ('herrmann', 0),
      ('scarlatti', 0),
      ('menditeguy', 0),
      ('gonzalez', 0),
      ('ireland', 0),
      ('thomson', 0),
      ('johnson', 0),
      ('ertl', 0),
      ('hartley', 0),
      ('stevenson', 0),
      ('freeland', 0),
      ('bettenhausen', 0),
      ('boyd', 0),
      ('gould', 0),
      ('behra', 0),
      ('paul_russo', 0),
      ('martini', 0),
      ('larini', 0),
      ('katayama', 0),
      ('morbidelli', 0),
      ('lamy', 0),
      ('brundle', 0),
      ('montermini', 0),
      ('blundell', 0),
      ('suzuki', 0),
      ('moreno', 0),
      ('wendlinger', 0),
      ('gachot', 0),
      ('magnussen', 0),
      ('tarquini', 0),
      ('comas', 0),
      ('bernard', 0),
      ('fittipaldi', 0),
      ('lehto', 0),
      ('cesaris', 0),
      ('alliot', 0),
      ('dalmas', 0),
      ('warwick', 0),
      ('capelli', 0),
      ('gugelmin', 0),
      ('rosa', 0),
      ('heidfeld', 0),
      ('kovalainen', 0),
      ('glock', 0),
      ('sato', 0),
      ('trulli', 0),
      ('sutil', 0),
      ('liuzzi', 0),
      ('wurz', 0),
      ('speed', 0),
      ('albers', 0),
      ('klien', 0),
      ('grouillard', 0),
      ('karthikeyan', 0),
      ('zonta', 0),
      ('gene', 0),
      ('verstappen', 0),
      ('alesi', 0),
      ('salo', 0),
      ('diniz', 0),
      ('buemi', 0),
      ('badoer', 0),
      ('zanardi', 0),
      ('hoffmann', 0),
      ('henton', 0),
      ('daly', 0),
      ('villota', 0),
      ('keegan', 0),
      ('rebaque', 0),
      ('merzario', 0),
      ('stuck', 0),
      ('lunger', 0),
      ('stommelen', 0),
      ('ian_scheckter', 0),
      ('pryce', 0),
      ('jarier', 0),
      ('oliver', 0),
      ('amon', 0),
      ('pescarolo', 0),
      ('wilson_fittipaldi', 0),
      ('keizan', 0),
      ('charlton', 0),
      ('hailwood', 0),
      ('ganley', 0),
      ('redman', 0),
      ('schenken', 0),
      ('palmer', 0),
      ('modena', 0),
      ('caffi', 0),
      ('lammers', 0),
      ('satoru_nakajima', 0),
      ('johansson', 0),
      ('nannini', 0),
      ('schneider', 0),
      ('giacomelli', 0),
      ('alguersuari', 0),
      ('grosjean', 0),
      ('kobayashi', 0),
      ('wisell', 0),
      ('danner', 0),
      ('cheever', 0),
      ('ghinzani', 0),
      ('streiff', 0),
      ('fabi', 0),
      ('surer', 0),
      ('manfred_winkelhock', 0),
      ('baldi', 0),
      ('serra', 0),
      ('salazar', 0);

ALTER TABLE "POPULATION__STAGING_TABLE_1" ADD COLUMN "t4__driverref__mapping_target_1_avg" REAL;

UPDATE "POPULATION__STAGING_TABLE_1" SET "t4__driverref__mapping_target_1_avg" = 0.0;

UPDATE "POPULATION__STAGING_TABLE_1"
SET "t4__driverref__mapping_target_1_avg" = "T4__DRIVERREF__MAPPING_TARGET_1_AVG"."value"
FROM "T4__DRIVERREF__MAPPING_TARGET_1_AVG"
WHERE "POPULATION__STAGING_TABLE_1"."t4__driverref" = "T4__DRIVERREF__MAPPING_TARGET_1_AVG"."key";

DROP TABLE IF EXISTS "T4__DRIVERREF__MAPPING_TARGET_1_AVG";

In [53]:

  Copied!     
 
pipe1.features.to_sql()[names[1]]
pipe1.features.to_sql()[names[1]]

Out[53]:

DROP TABLE IF EXISTS "FEATURE_1_75";

CREATE TABLE "FEATURE_1_75" AS
SELECT EWMA_90D( t2."win", t1."date" - t2."date__1_000000_days" ) AS "feature_1_75",
       t1.rowid AS rownum
FROM "POPULATION__STAGING_TABLE_1" t1
INNER JOIN "DRIVER_STANDINGS__STAGING_TABLE_2" t2
ON t1."driverid" = t2."driverid"
WHERE t2."date__1_000000_days" <= t1."date"
GROUP BY t1.rowid;

In [54]:

  Copied!     
 
pipe1.features.to_sql()[names[2]]
pipe1.features.to_sql()[names[2]]

Out[54]:

DROP TABLE IF EXISTS "FEATURE_1_78";

CREATE TABLE "FEATURE_1_78" AS
SELECT EWMA_365D( t2."win", t1."date" - t2."date__1_000000_days" ) AS "feature_1_78",
       t1.rowid AS rownum
FROM "POPULATION__STAGING_TABLE_1" t1
INNER JOIN "DRIVER_STANDINGS__STAGING_TABLE_2" t2
ON t1."driverid" = t2."driverid"
WHERE t2."date__1_000000_days" <= t1."date"
GROUP BY t1.rowid;

What we can learn from these features is the following: Knowing which driver we are talking aboyt and who won the most recent races is the best predictor for whether a driver will win this race.

In [55]:

  Copied!     
 
names, importances = pipe2.features.importances()

plt.subplots(figsize=(20, 10))

plt.bar(names[:30], importances[:30])

plt.title("feature importances")
plt.grid(True)
plt.xlabel("column")
plt.ylabel("importance")
plt.xticks(rotation='vertical')

plt.show()
names, importances = pipe2.features.importances() plt.subplots(figsize=(20, 10)) plt.bar(names[:30], importances[:30]) plt.title("feature importances") plt.grid(True) plt.xlabel("column") plt.ylabel("importance") plt.xticks(rotation='vertical') plt.show()

As we can see, featuretools generally builds features that are similar to the features identified by getML's FastProp.

2.7 Productionization¶

It is possible to productionize the pipeline by transpiling the features into production-ready SQL code. Please also refer to getML's sqlite3 and spark modules.

In [ ]:

  Copied!     
 
# Creates a folder named formula1_pipeline containing
# the SQL code.
pipe1.features.to_sql().save("formula1_pipeline")
# Creates a folder named formula1_pipeline containing # the SQL code. pipe1.features.to_sql().save("formula1_pipeline")

In [ ]:

  Copied!     
 
pipe1.features.to_sql(dialect=getml.pipeline.dialect.spark_sql).save("formula1_spark")
pipe1.features.to_sql(dialect=getml.pipeline.dialect.spark_sql).save("formula1_spark")

2.8 Discussion¶

For a more convenient overview, we summarize our results into a table.

In [58]:

  Copied!     
 
scores = [fastprop_score, featuretools_score]

pd.DataFrame(data={    'Name': ['getML: FastProp', 'featuretools'],
    'Accuracy': [f'{score.accuracy:.2%}' for score in scores],
    'AUC': [f'{score.auc:,.2%}' for score in scores],
    'Cross entropy': [f'{score.cross_entropy:,.5f}' for score in scores]
})
scores = [fastprop_score, featuretools_score] pd.DataFrame(data={ 'Name': ['getML: FastProp', 'featuretools'], 'Accuracy': [f'{score.accuracy:.2%}' for score in scores], 'AUC': [f'{score.auc:,.2%}' for score in scores], 'Cross entropy': [f'{score.cross_entropy:,.5f}' for score in scores] })

Out[58]:

	Name	Accuracy	AUC	Cross entropy
0	getML: FastProp	97.24%	92.35%	0.08533
1	featuretools	97.22%	91.53%	0.08797

As we can see, getML's FastProp outperforms featuretools according to all three measures.

In [ ]:

  Copied!     
 
getml.engine.shutdown()
getml.engine.shutdown()

3. Conclusion¶

We have benchmarked getML against featuretools on dataset related to Formula 1 races. We have found that getML's FastProp outperforms featuretools.

References¶

Motl, Jan, and Oliver Schulte. "The CTU prague relational learning repository." arXiv preprint arXiv:1511.03086 (2015).