Intégrer QueryData à une application

Ce tutoriel explique comment configurer et utiliser QueryData dans Bigtable à l'aide de la console Google Cloud , et comment l'intégrer à votre application. Découvrez comment créer le fichier d'ensemble de contextes, créer un ensemble de contextes qui utilise le fichier d'ensemble de contextes, utiliser MCP Toolbox pour appeler la méthode QueryData dans l'API Conversational Analytics afin de générer des requêtes SQL pour les questions en langage naturel, et l'intégrer à votre application.

Pour en savoir plus, consultez la présentation de QueryData.

Objectifs

Ce tutoriel vise les objectifs suivants :

  • Créer des tables de base de données et les remplir avec des données.
  • Créez un fichier d'ensemble de contextes avec l'agent d'ingénierie contextuelle.
  • Créez un ensemble de contextes et importez un fichier d'ensemble de contextes.
  • Testez QueryData et générez des requêtes SQL dans Bigtable Studio.
  • Intégrez QueryData à votre application à l'aide de l'outil Gemini Data Analytics QueryData dans MCP Toolbox.
  • Ajoutez l'ancrage aux réponses du LLM à l'aide de requêtes de recherche de valeurs.

Coûts

Dans ce document, vous utilisez les composants facturables suivants de Google Cloud :

Obtenez une estimation des coûts en fonction de votre utilisation prévue à l'aide du simulateur de coût.

Les nouveaux utilisateurs de Google Cloud peuvent bénéficier d'un essai sans frais.

Pour éviter de continuer à payer des frais, supprimez les ressources que vous avez créées lorsque vous avez terminé les tâches décrites dans ce document. Pour en savoir plus, consultez la section Effectuer un nettoyage.

Avant de commencer

Avant de créer un ensemble de contextes, assurez-vous de remplir les conditions préalables suivantes.

Activer les services requis

Activez les services suivants pour votre projet :

Préparer une instance Bigtable

Assurez-vous qu'une instance Bigtable est disponible. Pour en savoir plus, consultez la page Créer une instance.

Installer la CLI cbt

Installez et configurez la CLI cbt pour créer des tables et les remplir de données. Pour en savoir plus, consultez la présentation de la CLI cbt.

Rôles et autorisations nécessaires

  • Attribuez le rôle bigtable.admin à l'utilisateur IAM au niveau de l'instance ou du projet. Pour en savoir plus, consultez Contrôle des accès Bigtable avec IAM.
  • Attribuez le rôle geminidataanalytics.queryDataUser à l'utilisateur IAM au niveau du projet. Pour en savoir plus, consultez Ajouter une liaison de stratégie IAM pour un projet.
  • Pour créer, importer et gérer des ensembles de contextes, attribuez le rôle Éditeur de catalogue Dataplex (roles/dataplex.catalogEditor) ou Administrateur de catalogue Dataplex (roles/dataplex.catalogAdmin) à l'utilisateur IAM au niveau du projet. Pour en savoir plus, consultez Autorisations des ensembles de contextes.

Créer le schéma et les tables flights et airports

Dans cette section, vous allez créer les tables de base de données flights et airports pour ce tutoriel.

  1. Configurez votre projet et votre instance par défaut dans ~/.cbtrc : 
    echo project = PROJECT_ID >> ~/.cbtrc
echo instance = INSTANCE_ID >> ~/.cbtrc
     où :
    • PROJECT_ID : ID de votre projet Google Cloud .
    • INSTANCE_ID : ID de votre instance Bigtable.

  2. Exécutez la commande suivante pour créer la table airports et la famille de colonnes info :

    cbt createtable airports "families=info"

  3. Exécutez la commande suivante pour créer la table flights et la famille de colonnes info :

    cbt createtable flights "families=info"

Remplir les tables flights et airports

Dans cette section, vous allez remplir les tables flights et airports à l'aide de la commande cbt import.

  1. Créez un fichier nommé airports.csv avec le contenu suivant :

    ,info,info,info,info
,iata,name,city,country
AIRPORT#3249,BOS,General Edward Lawrence Logan International Airport,Boston,United States
AIRPORT#3257,OGG,Kahului Airport,Kahului,United States
AIRPORT#3259,MCI,Kansas City International Airport,Kansas City,United States
AIRPORT#3263,PHX,Phoenix Sky Harbor International Airport,Phoenix,United States
AIRPORT#3270,SFO,San Francisco International Airport,San Francisco,United States
AIRPORT#3285,LAX,Los Angeles International Airport,Los Angeles,United States
AIRPORT#3287,CLE,Cleveland Hopkins International Airport,Cleveland,United States
AIRPORT#3295,EWR,Newark Liberty International Airport,Newark,United States
AIRPORT#3296,BOI,Boise Air Terminal/Gowen Field,Boise,United States
AIRPORT#3303,DAL,Dallas Love Field,Dallas,United States
AIRPORT#3315,KOA,Ellison Onizuka Kona International At Keahole Airport,Kona,United States
AIRPORT#3321,DCA,Ronald Reagan Washington National Airport,Washington,United States
AIRPORT#3336,SLC,Salt Lake City International Airport,Salt Lake City,United States
AIRPORT#3350,IAH,George Bush Intercontinental Houston Airport,Houston,United States
AIRPORT#3370,PIT,Pittsburgh International Airport,Pittsburgh,United States
AIRPORT#3376,MIA,Miami International Airport,Miami,United States
AIRPORT#3377,SEA,Seattle Tacoma International Airport,Seattle,United States
AIRPORT#3382,LGB,Long Beach /Daugherty Field/ Airport,Long Beach,United States
AIRPORT#3385,IND,Indianapolis International Airport,Indianapolis,United States
AIRPORT#3402,LIH,Lihue Airport,Lihue,United States
AIRPORT#3421,SAT,San Antonio International Airport,San Antonio,United States
AIRPORT#3426,RDU,Raleigh Durham International Airport,Raleigh-durham,United States
AIRPORT#3436,TUS,Tucson International Airport,Tucson,United States
AIRPORT#3444,BUR,Bob Hope Airport,Burbank,United States
AIRPORT#3445,DTW,Detroit Metropolitan Wayne County Airport,Detroit,United States
AIRPORT#3470,DFW,Dallas Fort Worth International Airport,Dallas-Fort Worth,United States
AIRPORT#3473,AUS,Austin Bergstrom International Airport,Austin,United States
AIRPORT#3478,STL,St Louis Lambert International Airport,St. Louis,United States
AIRPORT#3482,ATL,Hartsfield Jackson Atlanta International Airport,Atlanta,United States
AIRPORT#3487,FAT,Fresno Yosemite International Airport,Fresno,United States
AIRPORT#3514,IAD,Washington Dulles International Airport,Washington,United States
AIRPORT#3517,MKE,General Mitchell International Airport,Milwaukee,United States
AIRPORT#3520,PDX,Portland International Airport,Portland,United States
AIRPORT#3528,HNL,Daniel K Inouye International Airport,Honolulu,United States
AIRPORT#3531,SAN,San Diego International Airport,San Diego,United States
AIRPORT#3534,ONT,Ontario International Airport,Ontario,United States
AIRPORT#3547,MDW,Chicago Midway International Airport,Chicago,United States
AIRPORT#3551,DEN,Denver International Airport,Denver,United States
AIRPORT#3552,PHL,Philadelphia International Airport,Philadelphia,United States
AIRPORT#3597,JFK,John F Kennedy International Airport,New York,United States
AIRPORT#3607,RNO,Reno Tahoe International Airport,Reno,United States
AIRPORT#3617,SMF,Sacramento International Airport,Sacramento,United States
AIRPORT#3619,COS,City of Colorado Springs Municipal Airport,Colorado Springs,United States
AIRPORT#3639,PSP,Palm Springs International Airport,Palm Springs,United States
AIRPORT#3649,BWI,Baltimore/Washington International Thurgood Marshall Airport,Baltimore,United States
AIRPORT#3658,MSP,Minneapolis-St Paul International/Wold-Chamberlain Airport,Minneapolis,United States
AIRPORT#3661,MSY,Louis Armstrong New Orleans International Airport,New Orleans,United States
AIRPORT#3667,SNA,John Wayne Airport-Orange County Airport,Santa Ana,United States
AIRPORT#3676,CLT,Charlotte Douglas International Airport,Charlotte,United States
AIRPORT#3677,LAS,McCarran International Airport,Las Vegas,United States
AIRPORT#3678,MCO,Orlando International Airport,Orlando,United States
AIRPORT#3742,MRY,Monterey Peninsula Airport,Monterey,United States
AIRPORT#3743,SBA,Santa Barbara Municipal Airport,Santa Barbara,United States
AIRPORT#3875,RDD,Redding Municipal Airport,Redding,United States
AIRPORT#3876,EUG,Mahlon Sweet Field,Eugene,United States
AIRPORT#3878,MFR,Rogue Valley International Medford Airport,Medford,United States
AIRPORT#3880,RDM,Roberts Field,Redmond-Bend,United States
AIRPORT#4111,ACV,California Redwood Coast-Humboldt County Airport,Arcata CA,United States
AIRPORT#4419,CEC,Jack Mc Namara Field Airport,Crescent City,United States
AIRPORT#4460,SBP,San Luis County Regional Airport,San Luis Obispo,United States
AIRPORT#4768,PSC,Tri Cities Airport,Pasco,United States
AIRPORT#5088,SUN,Friedman Memorial Airport,Hailey,United States
AIRPORT#5305,ASE,Aspen-Pitkin Co/Sardy Field,Aspen,United States
AIRPORT#5362,MMH,Mammoth Yosemite Airport,Mammoth Lakes,United States

  2. Importez les données du fichier airports.csv dans la table airports :

    cbt import airports airports.csv

  3. Créez un fichier nommé flights.csv avec le contenu suivant :

    ,info,info,info,info,info,info,info,info,info,info
,airline,flight_number,departure_airport,arrival_airport,departure_time,arrival_time,departure_gate,arrival_gate,departure_city,arrival_city
FLIGHT#0000,UA,1532,SFO,DEN,2025-01-01 05:50:00,2025-01-01 09:23:00,E49,D6,San Francisco,Denver
FLIGHT#0001,UA,1158,SFO,ORD,2025-01-01 05:57:00,2025-01-01 12:13:00,C38,D30,San Francisco,Chicago
FLIGHT#0002,F9,1057,IAH,SFO,2025-01-01 05:52:00,2025-01-01 08:14:00,E33,E43,Houston,San Francisco
FLIGHT#0003,OO,6403,MFR,SFO,2025-01-01 05:57:00,2025-01-01 07:23:00,A6,E50,Medford,San Francisco
FLIGHT#0004,OO,5628,RNO,SFO,2025-01-01 06:19:00,2025-01-01 07:37:00,A37,B29,Reno,San Francisco
FLIGHT#0005,CY,922,SFO,LAX,2025-01-01 06:38:00,2025-01-01 07:49:00,D20,E45,San Francisco,Los Angeles
FLIGHT#0006,AA,338,LAX,SFO,2025-01-01 06:44:00,2025-01-01 08:03:00,E20,A45,Los Angeles,San Francisco
FLIGHT#0007,UA,1195,MCO,SFO,2025-01-01 12:28:00,2025-01-01 15:06:00,D22,B36,Orlando,San Francisco
FLIGHT#0008,OO,5632,SBA,SFO,2025-01-01 06:56:00,2025-01-01 08:39:00,E19,A49,Santa Barbara,San Francisco
FLIGHT#0009,UA,300,CLE,SFO,2025-01-01 07:15:00,2025-01-01 09:22:00,E8,C28,Cleveland,San Francisco
FLIGHT#0010,UA,706,SEA,SFO,2025-01-01 07:14:00,2025-01-01 09:07:00,A32,B26,Seattle,San Francisco
FLIGHT#0011,AA,167,MIA,SFO,2025-01-01 07:47:00,2025-01-01 10:51:00,A23,E4,Miami,San Francisco
FLIGHT#0012,OO,5441,SFO,SBP,2025-01-01 07:08:00,2025-01-01 08:12:00,E21,D24,San Francisco,San Luis Obispo
FLIGHT#0013,UA,616,SFO,ORD,2025-01-01 07:14:00,2025-01-01 13:24:00,A11,D8,San Francisco,Chicago
FLIGHT#0014,AA,24,SFO,JFK,2025-01-01 07:14:00,2025-01-01 15:46:00,D4,C15,San Francisco,New York
FLIGHT#0015,UA,72,HNL,SFO,2025-01-01 07:28:00,2025-01-01 14:22:00,B33,D49,Honolulu,San Francisco
FLIGHT#0016,AS,244,PDX,SFO,2025-01-01 07:24:00,2025-01-01 09:04:00,E46,B8,Portland,San Francisco
FLIGHT#0017,OO,6307,MSP,SFO,2025-01-01 07:29:00,2025-01-01 09:32:00,C49,C36,Minneapolis,San Francisco
FLIGHT#0018,OO,4975,STL,SFO,2025-01-01 07:31:00,2025-01-01 09:58:00,C43,E16,St. Louis,San Francisco
FLIGHT#0019,AA,23,DFW,SFO,2025-01-01 08:41:00,2025-01-01 10:34:00,E13,E19,Dallas-Fort Worth,San Francisco
FLIGHT#0020,B6,434,SFO,BOS,2025-01-01 08:00:00,2025-01-01 16:23:00,D16,E46,San Francisco,Boston
FLIGHT#0021,F9,657,DEN,SFO,2025-01-01 07:58:00,2025-01-01 09:13:00,E41,E28,Denver,San Francisco
FLIGHT#0022,OO,6464,COS,SFO,2025-01-01 07:56:00,2025-01-01 09:45:00,D49,E17,Colorado Springs,San Francisco
FLIGHT#0023,UA,1235,IAD,SFO,2025-01-01 08:28:00,2025-01-01 11:15:00,A22,D19,Washington,San Francisco
FLIGHT#0024,OO,6527,EUG,SFO,2025-01-01 07:54:00,2025-01-01 09:36:00,D9,B26,Eugene,San Francisco
FLIGHT#0025,AA,242,SFO,ORD,2025-01-01 08:18:00,2025-01-01 14:26:00,E30,C1,San Francisco,Chicago
FLIGHT#0026,UA,1739,SFO,EWR,2025-01-01 08:22:00,2025-01-01 16:42:00,C13,D45,San Francisco,Newark
FLIGHT#0027,B6,133,BOS,SFO,2025-01-01 08:34:00,2025-01-01 12:30:00,D49,C16,Boston,San Francisco
FLIGHT#0028,OO,6336,SFO,RNO,2025-01-01 08:32:00,2025-01-01 09:29:00,B45,A2,San Francisco,Reno
FLIGHT#0029,UA,567,IAD,SFO,2025-01-01 10:47:00,2025-01-01 13:39:00,C27,B24,Washington,San Francisco
FLIGHT#0030,OO,5223,SLC,SFO,2025-01-01 08:50:00,2025-01-01 09:55:00,C38,C11,Salt Lake City,San Francisco
FLIGHT#0031,US,1784,SFO,CLT,2025-01-01 08:47:00,2025-01-01 16:32:00,C42,C31,San Francisco,Charlotte
FLIGHT#0032,WN,1579,ATL,SFO,2025-01-01 08:48:00,2025-01-01 10:55:00,C47,B5,Atlanta,San Francisco
FLIGHT#0033,OO,6396,BOI,SFO,2025-01-01 08:56:00,2025-01-01 09:43:00,E20,A6,Boise,San Francisco
FLIGHT#0034,AA,2293,DFW,SFO,2025-01-01 08:57:00,2025-01-01 11:08:00,D41,C37,Dallas-Fort Worth,San Francisco
FLIGHT#0035,DL,1106,SFO,MSP,2025-01-01 09:06:00,2025-01-01 14:56:00,E38,B30,San Francisco,Minneapolis
FLIGHT#0036,DL,1631,SFO,ATL,2025-01-01 09:00:00,2025-01-01 16:26:00,B13,B45,San Francisco,Atlanta
FLIGHT#0037,OO,5427,SFO,SMF,2025-01-01 09:06:00,2025-01-01 10:00:00,C34,C25,San Francisco,Sacramento
FLIGHT#0038,CY,352,SFO,BOS,2025-01-01 09:25:00,2025-01-01 17:51:00,A46,E42,San Francisco,Boston
FLIGHT#0039,UA,1239,LAX,SFO,2025-01-01 09:59:00,2025-01-01 11:19:00,A35,C2,Los Angeles,San Francisco
FLIGHT#0040,UA,1285,SFO,IAH,2025-01-01 09:41:00,2025-01-01 15:20:00,A39,E10,San Francisco,Houston
FLIGHT#0041,CY,753,SEA,SFO,2025-01-01 09:58:00,2025-01-01 11:53:00,C40,E35,Seattle,San Francisco
FLIGHT#0042,CY,956,SFO,SAN,2025-01-01 09:57:00,2025-01-01 11:30:00,B10,E44,San Francisco,San Diego
FLIGHT#0043,OO,5196,PDX,SFO,2025-01-01 09:53:00,2025-01-01 11:39:00,D13,C34,Portland,San Francisco
FLIGHT#0044,WN,2346,SFO,LAX,2025-01-01 09:59:00,2025-01-01 11:17:00,C24,E45,San Francisco,Los Angeles
FLIGHT#0045,OO,6400,EUG,SFO,2025-01-01 10:06:00,2025-01-01 11:27:00,E46,D20,Eugene,San Francisco
FLIGHT#0046,CY,23,JFK,SFO,2025-01-01 10:21:00,2025-01-01 14:09:00,D2,B21,New York,San Francisco
FLIGHT#0047,B6,534,SFO,BOS,2025-01-01 10:33:00,2025-01-01 18:55:00,A49,A5,San Francisco,Boston
FLIGHT#0048,UA,1727,BOS,SFO,2025-01-01 10:34:00,2025-01-01 13:57:00,D43,A31,Boston,San Francisco
FLIGHT#0049,WN,4391,LAS,SFO,2025-01-01 10:49:00,2025-01-01 12:17:00,A40,C49,Las Vegas,San Francisco
FLIGHT#0050,OO,5335,CEC,SFO,2025-01-01 10:25:00,2025-01-01 12:04:00,C8,C6,Crescent City,San Francisco
FLIGHT#0051,DL,1680,ATL,SFO,2025-01-01 10:43:00,2025-01-01 12:45:00,E49,E24,Atlanta,San Francisco
FLIGHT#0052,OO,6283,SFO,MCI,2025-01-01 10:38:00,2025-01-01 17:01:00,C32,B12,San Francisco,Kansas City
FLIGHT#0053,OO,5260,SFO,FAT,2025-01-01 10:37:00,2025-01-01 11:39:00,B46,B43,San Francisco,Fresno
FLIGHT#0054,OO,6249,LAS,SFO,2025-01-01 10:48:00,2025-01-01 12:19:00,D11,A49,Las Vegas,San Francisco
FLIGHT#0055,AA,23,SFO,DFW,2025-01-01 11:23:00,2025-01-01 17:00:00,B31,A11,San Francisco,Dallas-Fort Worth
FLIGHT#0056,UA,753,SFO,AUS,2025-01-01 10:59:00,2025-01-01 16:44:00,B18,D6,San Francisco,Austin
FLIGHT#0057,WN,4855,SFO,MKE,2025-01-01 11:08:00,2025-01-01 17:33:00,D3,D40,San Francisco,Milwaukee
FLIGHT#0058,DL,1845,SLC,SFO,2025-01-01 11:29:00,2025-01-01 12:06:00,A22,D2,Salt Lake City,San Francisco
FLIGHT#0059,UA,720,SFO,DEN,2025-01-01 12:46:00,2025-01-01 16:19:00,C41,E39,San Francisco,Denver
FLIGHT#0060,UA,1257,IAH,SFO,2025-01-01 11:30:00,2025-01-01 13:19:00,A26,D2,Houston,San Francisco
FLIGHT#0061,WN,2509,SFO,LAX,2025-01-01 11:34:00,2025-01-01 13:03:00,C42,D35,San Francisco,Los Angeles
FLIGHT#0062,UA,73,EWR,SFO,2025-01-01 12:11:00,2025-01-01 15:24:00,C5,E10,Newark,San Francisco
FLIGHT#0063,UA,364,PHX,SFO,2025-01-01 11:56:00,2025-01-01 13:11:00,A39,D26,Phoenix,San Francisco
FLIGHT#0064,AA,221,DFW,SFO,2025-01-01 11:55:00,2025-01-01 13:30:00,A47,B22,Dallas-Fort Worth,San Francisco
FLIGHT#0065,AA,2293,SFO,DFW,2025-01-01 12:04:00,2025-01-01 17:36:00,C43,E30,San Francisco,Dallas-Fort Worth
FLIGHT#0066,OO,5199,SFO,SLC,2025-01-01 12:09:00,2025-01-01 14:58:00,A40,E42,San Francisco,Salt Lake City
FLIGHT#0067,OO,6280,SLC,SFO,2025-01-01 12:13:00,2025-01-01 13:09:00,E16,E36,Salt Lake City,San Francisco
FLIGHT#0068,OO,6317,RNO,SFO,2025-01-01 12:05:00,2025-01-01 13:09:00,C23,A31,Reno,San Francisco
FLIGHT#0069,UA,864,SFO,EWR,2025-01-01 12:14:00,2025-01-01 20:38:00,C30,E37,San Francisco,Newark
FLIGHT#0070,OO,5400,ONT,SFO,2025-01-01 12:25:00,2025-01-01 13:40:00,A48,E19,Ontario,San Francisco
FLIGHT#0071,UA,236,HNL,SFO,2025-01-01 12:40:00,2025-01-01 19:36:00,A34,C12,Honolulu,San Francisco
FLIGHT#0072,OO,6343,SNA,SFO,2025-01-01 12:37:00,2025-01-01 13:57:00,A15,D26,Santa Ana,San Francisco
FLIGHT#0073,UA,1148,SFO,IAD,2025-01-01 13:10:00,2025-01-01 21:02:00,D24,A23,San Francisco,Washington
FLIGHT#0074,DL,432,JFK,SFO,2025-01-01 13:00:00,2025-01-01 16:28:00,E3,A35,New York,San Francisco
FLIGHT#0075,B6,1435,SFO,LGB,2025-01-01 13:33:00,2025-01-01 14:47:00,B24,C12,San Francisco,Long Beach
FLIGHT#0076,OO,6408,SFO,EUG,2025-01-01 13:43:00,2025-01-01 15:10:00,B47,A35,San Francisco,Eugene
FLIGHT#0077,UA,645,PDX,SFO,2025-01-01 13:27:00,2025-01-01 15:02:00,D7,E4,Portland,San Francisco
FLIGHT#0078,WN,1812,SFO,SAN,2025-01-01 14:32:00,2025-01-01 15:51:00,B35,D48,San Francisco,San Diego
FLIGHT#0079,OO,6409,SFO,PDX,2025-01-01 13:42:00,2025-01-01 15:27:00,C10,B17,San Francisco,Portland
FLIGHT#0080,UA,358,AUS,SFO,2025-01-01 14:31:00,2025-01-01 15:48:00,E33,D21,Austin,San Francisco
FLIGHT#0081,UA,1728,OGG,SFO,2025-01-01 14:02:00,2025-01-01 20:38:00,A30,C29,Kahului,San Francisco
FLIGHT#0082,UA,816,SEA,SFO,2025-01-01 13:44:00,2025-01-01 15:38:00,A36,A9,Seattle,San Francisco
FLIGHT#0083,UA,724,HNL,SFO,2025-01-01 13:59:00,2025-01-01 20:44:00,A14,C9,Honolulu,San Francisco
FLIGHT#0084,UA,1733,SFO,DEN,2025-01-01 14:38:00,2025-01-01 18:35:00,B27,E32,San Francisco,Denver
FLIGHT#0085,OO,5543,SFO,BOI,2025-01-01 14:27:00,2025-01-01 17:10:00,E10,B16,San Francisco,Boise
FLIGHT#0086,UA,1761,IAH,SFO,2025-01-01 14:55:00,2025-01-01 16:45:00,B2,E20,Houston,San Francisco
FLIGHT#0087,UA,277,DEN,SFO,2025-01-01 14:44:00,2025-01-01 16:04:00,B47,B13,Denver,San Francisco
FLIGHT#0088,AA,1213,DFW,SFO,2025-01-01 15:14:00,2025-01-01 16:45:00,D11,C31,Dallas-Fort Worth,San Francisco
FLIGHT#0089,US,407,PHX,SFO,2025-01-01 15:03:00,2025-01-01 16:00:00,C7,D19,Phoenix,San Francisco
FLIGHT#0090,OO,5367,SFO,RDM,2025-01-01 14:52:00,2025-01-01 16:23:00,A5,E40,San Francisco,Redmond-Bend
FLIGHT#0091,OO,5454,SMF,SFO,2025-01-01 15:46:00,2025-01-01 16:33:00,A20,A40,Sacramento,San Francisco
FLIGHT#0092,WN,2687,SFO,LAX,2025-01-01 15:15:00,2025-01-01 16:35:00,D41,B37,San Francisco,Los Angeles
FLIGHT#0093,CY,932,SFO,LAX,2025-01-01 15:31:00,2025-01-01 16:51:00,A19,E46,San Francisco,Los Angeles
FLIGHT#0094,WN,3236,SFO,SAN,2025-01-01 15:56:00,2025-01-01 17:23:00,B29,B32,San Francisco,San Diego
FLIGHT#0095,UA,404,IAH,SFO,2025-01-01 15:56:00,2025-01-01 17:52:00,B15,B42,Houston,San Francisco
FLIGHT#0096,CY,933,LAX,SFO,2025-01-01 15:42:00,2025-01-01 16:51:00,E16,E7,Los Angeles,San Francisco
FLIGHT#0097,US,596,SFO,PHX,2025-01-01 15:51:00,2025-01-01 18:52:00,B25,E38,San Francisco,Phoenix
FLIGHT#0098,AA,379,SFO,DFW,2025-01-01 16:04:00,2025-01-01 21:55:00,A30,D49,San Francisco,Dallas-Fort Worth
FLIGHT#0099,OO,6280,SFO,SNA,2025-01-01 16:52:00,2025-01-01 18:19:00,B5,C40,San Francisco,Santa Ana
FLIGHT#0100,UA,735,DEN,SFO,2025-01-01 17:34:00,2025-01-01 19:09:00,C31,B15,Denver,San Francisco
FLIGHT#0101,OO,5553,PDX,SFO,2025-01-01 16:00:00,2025-01-01 17:36:00,E9,D27,Portland,San Francisco
FLIGHT#0102,UA,73,SFO,HNL,2025-01-01 16:13:00,2025-01-01 19:52:00,C37,B37,San Francisco,Honolulu
FLIGHT#0103,UA,587,SEA,SFO,2025-01-01 16:25:00,2025-01-01 18:27:00,B7,B23,Seattle,San Francisco
FLIGHT#0104,OO,5363,SFO,ONT,2025-01-01 16:09:00,2025-01-01 17:28:00,D21,E29,San Francisco,Ontario
FLIGHT#0105,OO,6386,SFO,BOI,2025-01-01 16:12:00,2025-01-01 18:58:00,D48,E33,San Francisco,Boise
FLIGHT#0106,CY,27,JFK,SFO,2025-01-01 16:28:00,2025-01-01 20:19:00,C6,E14,New York,San Francisco
FLIGHT#0107,OO,5190,SFO,ASE,2025-01-01 16:24:00,2025-01-01 19:42:00,C50,B44,San Francisco,Aspen
FLIGHT#0108,CY,746,SFO,SEA,2025-01-01 16:40:00,2025-01-01 18:32:00,D35,C24,San Francisco,Seattle
FLIGHT#0109,UA,1640,SFO,ORD,2025-01-01 17:01:00,2025-01-01 23:02:00,E27,C24,San Francisco,Chicago
FLIGHT#0110,OO,6221,MCI,SFO,2025-01-01 17:28:00,2025-01-01 19:08:00,E2,C36,Kansas City,San Francisco
FLIGHT#0111,UA,486,SFO,LAS,2025-01-01 16:47:00,2025-01-01 18:16:00,C42,B47,San Francisco,Las Vegas
FLIGHT#0112,CY,2,DCA,SFO,2025-01-01 17:01:00,2025-01-01 20:07:00,E43,A33,Washington,San Francisco
FLIGHT#0113,B6,1436,LGB,SFO,2025-01-01 16:59:00,2025-01-01 18:35:00,B12,B22,Long Beach,San Francisco
FLIGHT#0114,WN,694,PHX,SFO,2025-01-01 17:11:00,2025-01-01 18:05:00,B44,B18,Phoenix,San Francisco
FLIGHT#0115,OO,5635,MMH,SFO,2025-01-01 17:00:00,2025-01-01 18:01:00,A25,D7,Mammoth Lakes,San Francisco
FLIGHT#0116,AS,341,PSP,SFO,2025-01-01 18:25:00,2025-01-01 19:46:00,A11,E18,Palm Springs,San Francisco
FLIGHT#0117,OO,6364,PSP,SFO,2025-01-01 18:01:00,2025-01-01 19:31:00,A14,B47,Palm Springs,San Francisco
FLIGHT#0118,OO,5418,SBP,SFO,2025-01-01 17:17:00,2025-01-01 18:15:00,B24,D17,San Luis Obispo,San Francisco
FLIGHT#0119,AA,197,SFO,ORD,2025-01-01 17:21:00,2025-01-01 23:33:00,D25,E49,San Francisco,Chicago
FLIGHT#0120,CY,720,SFO,DAL,2025-01-01 18:25:00,2025-01-01 23:43:00,B24,A2,San Francisco,Dallas
FLIGHT#0121,OO,6268,FAT,SFO,2025-01-01 17:33:00,2025-01-01 18:39:00,E35,D25,Fresno,San Francisco
FLIGHT#0122,OO,5615,SFO,EUG,2025-01-01 17:45:00,2025-01-01 19:18:00,B11,A34,San Francisco,Eugene
FLIGHT#0123,CY,960,SFO,SAN,2025-01-01 17:55:00,2025-01-01 19:23:00,D23,E44,San Francisco,San Diego
FLIGHT#0124,AA,1393,DFW,SFO,2025-01-01 17:51:00,2025-01-01 19:25:00,E11,A41,Dallas-Fort Worth,San Francisco
FLIGHT#0125,OO,6352,BOI,SFO,2025-01-01 17:58:00,2025-01-01 18:46:00,B31,D23,Boise,San Francisco
FLIGHT#0126,DL,1151,SFO,HNL,2025-01-01 17:53:00,2025-01-01 21:34:00,E29,B45,San Francisco,Honolulu
FLIGHT#0127,OO,5227,SFO,SAN,2025-01-01 19:07:00,2025-01-01 20:31:00,A35,B38,San Francisco,San Diego
FLIGHT#0128,CY,941,LAX,SFO,2025-01-01 18:19:00,2025-01-01 19:34:00,C33,B44,Los Angeles,San Francisco
FLIGHT#0129,UA,1044,EWR,SFO,2025-01-01 19:19:00,2025-01-01 22:53:00,D50,D26,Newark,San Francisco
FLIGHT#0130,CY,211,ORD,SFO,2025-01-01 18:41:00,2025-01-01 21:14:00,D13,E45,Chicago,San Francisco
FLIGHT#0131,UA,685,ORD,SFO,2025-01-01 18:51:00,2025-01-01 21:04:00,D26,D45,Chicago,San Francisco
FLIGHT#0132,UA,370,SFO,SEA,2025-01-01 18:49:00,2025-01-01 20:41:00,B6,E20,San Francisco,Seattle
FLIGHT#0133,UA,1433,IAH,SFO,2025-01-01 19:45:00,2025-01-01 21:39:00,C42,C48,Houston,San Francisco
FLIGHT#0134,UA,1172,DEN,SFO,2025-01-01 20:19:00,2025-01-01 22:37:00,E31,D43,Denver,San Francisco
FLIGHT#0135,CY,593,PSP,SFO,2025-01-01 19:09:00,2025-01-01 20:26:00,C23,C43,Palm Springs,San Francisco
FLIGHT#0136,UA,1564,SFO,ORD,2025-01-01 19:14:00,2025-01-02 01:14:00,E3,C48,San Francisco,Chicago
FLIGHT#0137,UA,741,ORD,SFO,2025-01-01 19:18:00,2025-01-01 21:58:00,A25,B38,Chicago,San Francisco
FLIGHT#0138,WN,600,SFO,SNA,2025-01-01 19:36:00,2025-01-01 20:53:00,D21,D11,San Francisco,Santa Ana
FLIGHT#0139,F9,664,SFO,DEN,2025-01-01 19:31:00,2025-01-01 23:13:00,E11,E12,San Francisco,Denver
FLIGHT#0140,OO,4995,LAS,SFO,2025-01-01 19:18:00,2025-01-01 20:58:00,B42,B49,Las Vegas,San Francisco
FLIGHT#0141,CY,717,DAL,SFO,2025-01-01 19:31:00,2025-01-01 21:02:00,A42,D32,Dallas,San Francisco
FLIGHT#0142,UA,1243,SFO,DEN,2025-01-01 19:38:00,2025-01-01 23:15:00,B34,A3,San Francisco,Denver
FLIGHT#0143,UA,560,IAD,SFO,2025-01-01 19:41:00,2025-01-01 22:29:00,D9,D33,Washington,San Francisco
FLIGHT#0144,WN,2791,LAX,SFO,2025-01-01 19:45:00,2025-01-01 20:57:00,C49,C22,Los Angeles,San Francisco
FLIGHT#0145,WN,2079,SAN,SFO,2025-01-01 21:03:00,2025-01-01 22:33:00,A25,E24,San Diego,San Francisco
FLIGHT#0146,OO,6356,SFO,SLC,2025-01-01 19:57:00,2025-01-01 22:49:00,A28,C45,San Francisco,Salt Lake City
FLIGHT#0147,OO,6290,PSP,SFO,2025-01-01 19:48:00,2025-01-01 21:12:00,A33,E27,Palm Springs,San Francisco
FLIGHT#0148,OO,5363,ONT,SFO,2025-01-01 20:27:00,2025-01-01 21:54:00,A14,E17,Ontario,San Francisco
FLIGHT#0149,OO,5206,SFO,PHX,2025-01-01 20:12:00,2025-01-01 23:05:00,D22,E31,San Francisco,Phoenix
FLIGHT#0150,B6,1136,LGB,SFO,2025-01-01 20:41:00,2025-01-01 21:56:00,B7,E45,Long Beach,San Francisco
FLIGHT#0151,US,736,PHX,SFO,2025-01-01 21:12:00,2025-01-01 22:11:00,B22,E33,Phoenix,San Francisco
FLIGHT#0152,AA,1611,DFW,SFO,2025-01-01 21:42:00,2025-01-01 23:37:00,A13,B44,Dallas-Fort Worth,San Francisco
FLIGHT#0153,UA,1241,SFO,EWR,2025-01-01 21:44:00,2025-01-02 05:55:00,D3,A17,San Francisco,Newark
FLIGHT#0154,UA,1726,OGG,SFO,2025-01-01 22:19:00,2025-01-02 04:56:00,C35,B23,Kahului,San Francisco
FLIGHT#0155,US,898,SFO,CLT,2025-01-01 22:29:00,2025-01-02 06:18:00,D7,D15,San Francisco,Charlotte
FLIGHT#0156,UA,1669,SFO,PHL,2025-01-01 22:26:00,2025-01-02 06:30:00,B13,A12,San Francisco,Philadelphia
FLIGHT#0157,UA,362,SFO,BWI,2025-01-01 22:31:00,2025-01-02 06:36:00,E32,A36,San Francisco,Baltimore
FLIGHT#0158,OO,6220,SFO,RNO,2025-01-01 22:25:00,2025-01-01 23:23:00,D18,A31,San Francisco,Reno
FLIGHT#0159,OO,6376,SFO,MRY,2025-01-01 22:35:00,2025-01-01 23:19:00,C24,E48,San Francisco,Monterey
FLIGHT#0160,UA,522,SFO,PDX,2025-01-01 22:36:00,2025-01-02 00:31:00,B17,A40,San Francisco,Portland
FLIGHT#0161,UA,1728,SFO,BOS,2025-01-01 22:46:00,2025-01-02 07:20:00,C40,E43,San Francisco,Boston
FLIGHT#0162,UA,724,SFO,LAX,2025-01-01 22:45:00,2025-01-02 00:29:00,E43,B23,San Francisco,Los Angeles
FLIGHT#0163,UA,1681,SFO,CLE,2025-01-01 23:03:00,2025-01-02 06:45:00,D26,E19,San Francisco,Cleveland
FLIGHT#0164,UA,1580,OGG,SFO,2025-01-01 23:24:00,2025-01-02 06:07:00,E38,A35,Kahului,San Francisco
FLIGHT#0165,DL,806,SFO,MSP,2025-01-02 00:22:00,2025-01-02 05:57:00,B27,A3,San Francisco,Minneapolis
FLIGHT#0166,OO,5467,ONT,SFO,2025-01-02 05:02:00,2025-01-02 06:18:00,E8,C32,Ontario,San Francisco
FLIGHT#0167,WN,1989,SFO,LAS,2025-01-02 05:35:00,2025-01-02 06:46:00,A21,D15,San Francisco,Las Vegas
FLIGHT#0168,UA,1619,BOS,SFO,2025-01-02 05:56:00,2025-01-02 09:56:00,C46,E11,Boston,San Francisco
FLIGHT#0169,UA,1076,SFO,ORD,2025-01-02 05:58:00,2025-01-02 12:00:00,E27,A31,San Francisco,Chicago
FLIGHT#0170,F9,1056,SFO,IAH,2025-01-02 05:59:00,2025-01-02 12:09:00,A29,B13,San Francisco,Houston
FLIGHT#0171,WN,2242,LAX,SFO,2025-01-02 06:22:00,2025-01-02 07:36:00,C12,D43,Los Angeles,San Francisco
FLIGHT#0172,B6,735,SFO,LGB,2025-01-02 06:01:00,2025-01-02 07:20:00,C2,C37,San Francisco,Long Beach
FLIGHT#0173,WN,3859,SFO,MDW,2025-01-02 06:27:00,2025-01-02 12:26:00,A28,E38,San Francisco,Chicago
FLIGHT#0174,UA,1285,SFO,ORD,2025-01-02 06:43:00,2025-01-02 12:47:00,B43,A35,San Francisco,Chicago
FLIGHT#0175,WN,1726,SFO,PHX,2025-01-02 06:32:00,2025-01-02 09:20:00,C16,B2,San Francisco,Phoenix
FLIGHT#0176,OO,5186,SFO,SAN,2025-01-02 06:25:00,2025-01-02 07:59:00,C47,C1,San Francisco,San Diego
FLIGHT#0177,UA,566,PIT,SFO,2025-01-02 07:15:00,2025-01-02 09:46:00,C39,E39,Pittsburgh,San Francisco
FLIGHT#0178,UA,1195,MCO,SFO,2025-01-02 07:18:00,2025-01-02 10:05:00,A47,C43,Orlando,San Francisco
FLIGHT#0179,CY,925,LAX,SFO,2025-01-02 06:57:00,2025-01-02 08:07:00,E29,C19,Los Angeles,San Francisco
FLIGHT#0180,UA,761,ORD,SFO,2025-01-02 07:07:00,2025-01-02 09:34:00,B22,A3,Chicago,San Francisco
FLIGHT#0181,UA,758,SFO,JFK,2025-01-02 07:07:00,2025-01-02 15:53:00,B15,B39,San Francisco,New York
FLIGHT#0182,CY,67,IAD,SFO,2025-01-02 07:07:00,2025-01-02 10:00:00,D24,E33,Washington,San Francisco
FLIGHT#0183,UA,706,SEA,SFO,2025-01-02 07:05:00,2025-01-02 08:59:00,D9,C28,Seattle,San Francisco
FLIGHT#0184,OO,5441,SFO,SBP,2025-01-02 07:06:00,2025-01-02 08:11:00,B3,B29,San Francisco,San Luis Obispo
FLIGHT#0185,CY,202,SFO,ORD,2025-01-02 07:18:00,2025-01-02 13:23:00,C16,D10,San Francisco,Chicago
FLIGHT#0186,CY,710,SFO,DAL,2025-01-02 07:16:00,2025-01-02 12:43:00,D1,E13,San Francisco,Dallas
FLIGHT#0187,AA,24,SFO,JFK,2025-01-02 07:06:00,2025-01-02 15:29:00,C38,D2,San Francisco,New York
FLIGHT#0188,UA,72,HNL,SFO,2025-01-02 07:28:00,2025-01-02 14:20:00,C8,E18,Honolulu,San Francisco
FLIGHT#0189,UA,1278,ATL,SFO,2025-01-02 07:30:00,2025-01-02 09:41:00,C16,B21,Atlanta,San Francisco
FLIGHT#0190,UA,498,SFO,SAN,2025-01-02 07:20:00,2025-01-02 08:38:00,B42,E44,San Francisco,San Diego
FLIGHT#0191,CY,183,EWR,SFO,2025-01-02 07:26:00,2025-01-02 10:51:00,B33,A12,Newark,San Francisco
FLIGHT#0192,AA,76,SFO,LAX,2025-01-02 07:22:00,2025-01-02 08:44:00,E28,B18,San Francisco,Los Angeles
FLIGHT#0193,WN,2325,DEN,SFO,2025-01-02 07:40:00,2025-01-02 09:03:00,A5,A17,Denver,San Francisco
FLIGHT#0194,WN,1545,LAX,SFO,2025-01-02 07:48:00,2025-01-02 09:02:00,D42,A25,Los Angeles,San Francisco
FLIGHT#0195,OO,5196,SFO,PDX,2025-01-02 07:46:00,2025-01-02 09:32:00,B12,B11,San Francisco,Portland
FLIGHT#0196,UA,1053,DEN,SFO,2025-01-02 08:06:00,2025-01-02 09:41:00,C37,E7,Denver,San Francisco
FLIGHT#0197,AA,23,DFW,SFO,2025-01-02 07:53:00,2025-01-02 09:26:00,D24,C34,Dallas-Fort Worth,San Francisco
FLIGHT#0198,UA,322,SNA,SFO,2025-01-02 07:53:00,2025-01-02 09:17:00,E22,D32,Santa Ana,San Francisco
FLIGHT#0199,UA,500,IND,SFO,2025-01-02 08:02:00,2025-01-02 09:46:00,D47,B25,Indianapolis,San Francisco
FLIGHT#0200,UA,662,SAN,SFO,2025-01-02 08:51:00,2025-01-02 10:38:00,C36,B37,San Diego,San Francisco
FLIGHT#0201,UA,834,LAS,SFO,2025-01-02 07:51:00,2025-01-02 09:17:00,B24,E14,Las Vegas,San Francisco
FLIGHT#0202,CY,1930,SFO,LAX,2025-01-02 07:58:00,2025-01-02 09:29:00,D1,E49,San Francisco,Los Angeles
FLIGHT#0203,UA,1454,ORD,SFO,2025-01-02 08:09:00,2025-01-02 10:54:00,C45,D18,Chicago,San Francisco
FLIGHT#0204,AA,1368,SFO,DFW,2025-01-02 08:11:00,2025-01-02 13:35:00,A24,D28,San Francisco,Dallas-Fort Worth
FLIGHT#0205,UA,460,LAX,SFO,2025-01-02 08:22:00,2025-01-02 09:53:00,E4,D14,Los Angeles,San Francisco
FLIGHT#0206,UA,1547,SFO,BOS,2025-01-02 08:30:00,2025-01-02 16:52:00,B24,D37,San Francisco,Boston
FLIGHT#0207,AA,242,SFO,ORD,2025-01-02 08:28:00,2025-01-02 14:41:00,D10,B31,San Francisco,Chicago
FLIGHT#0208,OO,5538,MFR,SFO,2025-01-02 08:15:00,2025-01-02 09:44:00,D3,B43,Medford,San Francisco
FLIGHT#0209,UA,1739,SFO,EWR,2025-01-02 08:56:00,2025-01-02 17:03:00,B18,A15,San Francisco,Newark
FLIGHT#0210,DL,745,DTW,SFO,2025-01-02 08:36:00,2025-01-02 10:43:00,B2,D29,Detroit,San Francisco
FLIGHT#0211,OO,5445,SFO,SBP,2025-01-02 10:35:00,2025-01-02 11:50:00,D39,A1,San Francisco,San Luis Obispo
FLIGHT#0212,OO,5223,SLC,SFO,2025-01-02 08:52:00,2025-01-02 09:50:00,C17,C19,Salt Lake City,San Francisco
FLIGHT#0213,UA,1615,SFO,IAD,2025-01-02 08:44:00,2025-01-02 16:41:00,A19,C46,San Francisco,Washington
FLIGHT#0214,AA,1435,ORD,SFO,2025-01-02 08:43:00,2025-01-02 11:08:00,B49,B11,Chicago,San Francisco
FLIGHT#0215,HA,11,SFO,HNL,2025-01-02 09:28:00,2025-01-02 13:38:00,E28,B21,San Francisco,Honolulu
FLIGHT#0216,OO,6373,SFO,DFW,2025-01-02 08:51:00,2025-01-02 14:40:00,B42,D13,San Francisco,Dallas-Fort Worth
FLIGHT#0217,UA,479,SFO,LAX,2025-01-02 11:15:00,2025-01-02 12:47:00,E32,B7,San Francisco,Los Angeles
FLIGHT#0218,UA,361,SFO,LIH,2025-01-02 09:33:00,2025-01-02 13:10:00,A45,C26,San Francisco,Lihue
FLIGHT#0219,WN,2074,SFO,SNA,2025-01-02 09:08:00,2025-01-02 10:18:00,A6,A49,San Francisco,Santa Ana
FLIGHT#0220,CY,84,SFO,IAD,2025-01-02 09:20:00,2025-01-02 17:25:00,A37,B4,San Francisco,Washington
FLIGHT#0221,WN,2969,SAN,SFO,2025-01-02 09:35:00,2025-01-02 10:59:00,B50,A43,San Diego,San Francisco
FLIGHT#0222,UA,1249,SFO,MCO,2025-01-02 10:10:00,2025-01-02 18:15:00,D13,E20,San Francisco,Orlando
FLIGHT#0223,WN,1820,SFO,LAS,2025-01-02 09:52:00,2025-01-02 11:19:00,B19,D16,San Francisco,Las Vegas
FLIGHT#0224,WN,2414,SFO,PHX,2025-01-02 09:57:00,2025-01-02 12:52:00,E1,C5,San Francisco,Phoenix
FLIGHT#0225,UA,698,SEA,SFO,2025-01-02 09:54:00,2025-01-02 11:59:00,E37,D34,Seattle,San Francisco
FLIGHT#0226,AS,303,SFO,SEA,2025-01-02 10:06:00,2025-01-02 12:11:00,D36,C9,San Francisco,Seattle
FLIGHT#0227,WN,3323,SFO,DEN,2025-01-02 10:40:00,2025-01-02 13:59:00,A4,E35,San Francisco,Denver
FLIGHT#0228,OO,6286,SBA,SFO,2025-01-02 10:23:00,2025-01-02 11:47:00,B46,A8,Santa Barbara,San Francisco
FLIGHT#0229,OO,5603,PHX,SFO,2025-01-02 10:29:00,2025-01-02 11:41:00,A2,C22,Phoenix,San Francisco
FLIGHT#0230,CY,218,SFO,AUS,2025-01-02 10:28:00,2025-01-02 15:56:00,E49,C21,San Francisco,Austin
FLIGHT#0231,OO,5335,CEC,SFO,2025-01-02 10:32:00,2025-01-02 12:09:00,D29,D17,Crescent City,San Francisco
FLIGHT#0232,OO,5646,SFO,SBA,2025-01-02 10:33:00,2025-01-02 11:54:00,A50,E40,San Francisco,Santa Barbara
FLIGHT#0233,OO,5635,SFO,SNA,2025-01-02 11:22:00,2025-01-02 12:55:00,C17,D47,San Francisco,Santa Ana
FLIGHT#0234,OO,6380,SFO,BOI,2025-01-02 10:41:00,2025-01-02 13:18:00,B18,C15,San Francisco,Boise
FLIGHT#0235,UA,662,SFO,MSP,2025-01-02 10:52:00,2025-01-02 16:39:00,B6,E11,San Francisco,Minneapolis
FLIGHT#0236,OO,6270,SFO,RNO,2025-01-02 10:43:00,2025-01-02 11:52:00,C15,E10,San Francisco,Reno
FLIGHT#0237,OO,4975,SFO,TUS,2025-01-02 12:23:00,2025-01-02 15:18:00,D5,C12,San Francisco,Tucson
FLIGHT#0238,UA,322,SFO,SEA,2025-01-02 10:49:00,2025-01-02 12:54:00,E29,D3,San Francisco,Seattle
FLIGHT#0239,WN,1009,LAX,SFO,2025-01-02 11:33:00,2025-01-02 12:45:00,E12,C30,Los Angeles,San Francisco
FLIGHT#0240,OO,4966,SFO,SUN,2025-01-02 11:11:00,2025-01-02 14:15:00,D39,D21,San Francisco,Hailey
FLIGHT#0241,UA,753,SFO,AUS,2025-01-02 11:40:00,2025-01-02 17:33:00,B30,D47,San Francisco,Austin
FLIGHT#0242,UA,242,SFO,SAN,2025-01-02 12:00:00,2025-01-02 13:24:00,B27,E27,San Francisco,San Diego
FLIGHT#0243,UA,390,SEA,SFO,2025-01-02 11:10:00,2025-01-02 13:08:00,A17,A23,Seattle,San Francisco
FLIGHT#0244,OO,6336,RNO,SFO,2025-01-02 11:17:00,2025-01-02 12:17:00,A45,E39,Reno,San Francisco
FLIGHT#0245,OO,5250,MRY,SFO,2025-01-02 11:12:00,2025-01-02 12:28:00,B33,E11,Monterey,San Francisco
FLIGHT#0246,UA,317,SAN,SFO,2025-01-02 12:10:00,2025-01-02 13:40:00,A14,A42,San Diego,San Francisco
FLIGHT#0247,OO,5652,SFO,PSP,2025-01-02 12:14:00,2025-01-02 13:50:00,E17,B17,San Francisco,Palm Springs
FLIGHT#0248,UA,1195,SFO,ORD,2025-01-02 11:56:00,2025-01-02 18:12:00,B39,B19,San Francisco,Chicago
FLIGHT#0249,CY,714,SFO,DAL,2025-01-02 11:40:00,2025-01-02 17:22:00,C50,A19,San Francisco,Dallas
FLIGHT#0250,WN,2779,LAX,SFO,2025-01-02 11:58:00,2025-01-02 13:09:00,B19,C27,Los Angeles,San Francisco
FLIGHT#0251,UA,1722,SFO,IAH,2025-01-02 11:58:00,2025-01-02 18:29:00,E12,D6,San Francisco,Houston
FLIGHT#0252,UA,766,JFK,SFO,2025-01-02 11:55:00,2025-01-02 15:21:00,A16,C18,New York,San Francisco
FLIGHT#0253,UA,1437,ORD,SFO,2025-01-02 12:36:00,2025-01-02 15:09:00,A33,E19,Chicago,San Francisco
FLIGHT#0254,UA,1294,LAS,SFO,2025-01-02 12:20:00,2025-01-02 13:45:00,E7,A10,Las Vegas,San Francisco
FLIGHT#0255,UA,1266,SEA,SFO,2025-01-02 12:13:00,2025-01-02 14:11:00,B32,C18,Seattle,San Francisco
FLIGHT#0256,UA,1668,SFO,EWR,2025-01-02 12:24:00,2025-01-02 20:34:00,D37,A30,San Francisco,Newark
FLIGHT#0257,OO,6304,BUR,SFO,2025-01-02 12:31:00,2025-01-02 14:01:00,C7,C4,Burbank,San Francisco
FLIGHT#0258,WN,4391,SFO,STL,2025-01-02 12:55:00,2025-01-02 18:36:00,A12,D39,San Francisco,St. Louis
FLIGHT#0259,UA,236,HNL,SFO,2025-01-02 13:19:00,2025-01-02 20:12:00,B25,D26,Honolulu,San Francisco
FLIGHT#0260,UA,1705,SFO,IAH,2025-01-02 12:51:00,2025-01-02 18:50:00,E11,E32,San Francisco,Houston
FLIGHT#0261,OO,6490,EUG,SFO,2025-01-02 13:00:00,2025-01-02 14:21:00,D34,D29,Eugene,San Francisco
FLIGHT#0262,UA,820,SFO,ORD,2025-01-02 13:02:00,2025-01-02 19:11:00,B45,E24,San Francisco,Chicago
FLIGHT#0263,AS,381,SFO,PDX,2025-01-02 12:52:00,2025-01-02 14:34:00,E11,B3,San Francisco,Portland
FLIGHT#0264,AS,305,SFO,SEA,2025-01-02 12:59:00,2025-01-02 15:10:00,C34,C41,San Francisco,Seattle
FLIGHT#0265,OO,6273,SFO,SLC,2025-01-02 12:57:00,2025-01-02 15:50:00,C3,D36,San Francisco,Salt Lake City
FLIGHT#0266,DL,2254,SFO,DTW,2025-01-02 13:05:00,2025-01-02 20:29:00,E10,D30,San Francisco,Detroit
FLIGHT#0267,B6,1435,SFO,LGB,2025-01-02 13:34:00,2025-01-02 15:02:00,B31,D48,San Francisco,Long Beach
FLIGHT#0268,F9,661,DEN,SFO,2025-01-02 13:11:00,2025-01-02 14:35:00,C3,E38,Denver,San Francisco
FLIGHT#0269,UA,1092,EWR,SFO,2025-01-02 13:44:00,2025-01-02 17:05:00,D13,A41,Newark,San Francisco
FLIGHT#0270,US,1812,SFO,CLT,2025-01-02 13:32:00,2025-01-02 21:36:00,A50,A5,San Francisco,Charlotte
FLIGHT#0271,WN,2598,LAX,SFO,2025-01-02 13:48:00,2025-01-02 14:54:00,E23,C38,Los Angeles,San Francisco
FLIGHT#0272,UA,216,SFO,SNA,2025-01-02 15:23:00,2025-01-02 16:41:00,C23,E6,San Francisco,Santa Ana
FLIGHT#0273,UA,1060,SFO,DCA,2025-01-02 15:19:00,2025-01-02 23:18:00,D14,B18,San Francisco,Washington
FLIGHT#0274,OO,6326,BOI,SFO,2025-01-02 13:51:00,2025-01-02 14:24:00,D49,A43,Boise,San Francisco
FLIGHT#0275,UA,724,HNL,SFO,2025-01-02 14:43:00,2025-01-02 21:33:00,B26,A50,Honolulu,San Francisco
FLIGHT#0276,OO,5463,SFO,SMF,2025-01-02 13:50:00,2025-01-02 14:32:00,A13,D23,San Francisco,Sacramento
FLIGHT#0277,UA,698,SFO,MCO,2025-01-02 14:24:00,2025-01-02 22:33:00,D49,B2,San Francisco,Orlando
FLIGHT#0278,CY,1178,SFO,EWR,2025-01-02 13:58:00,2025-01-02 22:06:00,C27,D9,San Francisco,Newark
FLIGHT#0279,OO,6425,TUS,SFO,2025-01-02 14:50:00,2025-01-02 15:54:00,E29,D17,Tucson,San Francisco
FLIGHT#0280,OO,5615,MFR,SFO,2025-01-02 13:55:00,2025-01-02 15:15:00,B44,E27,Medford,San Francisco
FLIGHT#0281,WN,3450,SFO,MDW,2025-01-02 14:50:00,2025-01-02 20:46:00,D39,B44,San Francisco,Chicago
FLIGHT#0282,UA,1733,SFO,DEN,2025-01-02 15:11:00,2025-01-02 18:47:00,D8,D2,San Francisco,Denver
FLIGHT#0283,UA,1263,KOA,SFO,2025-01-02 14:59:00,2025-01-02 21:49:00,C39,E3,Kona,San Francisco
FLIGHT#0284,UA,1568,SFO,RDU,2025-01-02 14:55:00,2025-01-02 23:03:00,A30,A26,San Francisco,Raleigh-durham
FLIGHT#0285,UA,1159,SFO,BOS,2025-01-02 14:57:00,2025-01-02 23:23:00,C19,A30,San Francisco,Boston
FLIGHT#0286,OO,5652,PSP,SFO,2025-01-02 14:52:00,2025-01-02 16:46:00,D37,A5,Palm Springs,San Francisco
FLIGHT#0287,US,749,CLT,SFO,2025-01-02 15:28:00,2025-01-02 17:47:00,D17,D38,Charlotte,San Francisco
FLIGHT#0288,UA,1491,SFO,DEN,2025-01-02 15:09:00,2025-01-02 18:45:00,E31,E36,San Francisco,Denver
FLIGHT#0289,WN,2687,SFO,LAX,2025-01-02 15:27:00,2025-01-02 16:31:00,E42,B13,San Francisco,Los Angeles
FLIGHT#0290,DL,2040,SFO,JFK,2025-01-02 15:09:00,2025-01-02 23:20:00,E2,D33,San Francisco,New York
FLIGHT#0291,OO,5549,SFO,AUS,2025-01-02 15:24:00,2025-01-02 21:11:00,B42,C1,San Francisco,Austin
FLIGHT#0292,AS,308,SEA,SFO,2025-01-02 15:22:00,2025-01-02 17:18:00,C39,A46,Seattle,San Francisco
FLIGHT#0293,F9,662,SFO,DEN,2025-01-02 15:16:00,2025-01-02 18:50:00,C5,A8,San Francisco,Denver
FLIGHT#0294,UA,735,DEN,SFO,2025-01-02 16:00:00,2025-01-02 17:19:00,A32,C24,Denver,San Francisco
FLIGHT#0295,UA,404,IAH,SFO,2025-01-02 17:26:00,2025-01-02 19:32:00,C14,A7,Houston,San Francisco
FLIGHT#0296,CY,210,SFO,ORD,2025-01-02 15:42:00,2025-01-02 21:44:00,A31,B46,San Francisco,Chicago
FLIGHT#0297,AS,315,SFO,SEA,2025-01-02 19:47:00,2025-01-02 22:16:00,A8,B49,San Francisco,Seattle
FLIGHT#0298,WN,1901,SAN,SFO,2025-01-02 15:48:00,2025-01-02 17:09:00,A23,B38,San Diego,San Francisco
FLIGHT#0299,OO,5580,SFO,SNA,2025-01-02 16:35:00,2025-01-02 18:03:00,D40,E43,San Francisco,Santa Ana
FLIGHT#0300,AA,2265,DFW,SFO,2025-01-02 16:07:00,2025-01-02 17:41:00,A4,A23,Dallas-Fort Worth,San Francisco
FLIGHT#0301,OO,5363,SFO,ONT,2025-01-02 16:15:00,2025-01-02 17:32:00,A10,A26,San Francisco,Ontario
FLIGHT#0302,OO,5609,SFO,PDX,2025-01-02 17:01:00,2025-01-02 18:42:00,E49,D4,San Francisco,Portland
FLIGHT#0303,OO,5645,MFR,SFO,2025-01-02 16:15:00,2025-01-02 17:35:00,A25,E49,Medford,San Francisco
FLIGHT#0304,OO,4976,SFO,STL,2025-01-02 16:37:00,2025-01-02 22:27:00,C35,A2,San Francisco,St. Louis
FLIGHT#0305,OO,6217,BUR,SFO,2025-01-02 16:44:00,2025-01-02 17:55:00,D50,A14,Burbank,San Francisco
FLIGHT#0306,WN,1276,SFO,LAX,2025-01-02 16:25:00,2025-01-02 17:32:00,B8,D22,San Francisco,Los Angeles
FLIGHT#0307,OO,6238,ACV,SFO,2025-01-02 16:25:00,2025-01-02 17:36:00,E24,A39,Arcata CA,San Francisco
FLIGHT#0308,OO,5190,SFO,ASE,2025-01-02 17:08:00,2025-01-02 20:17:00,C9,E6,San Francisco,Aspen
FLIGHT#0309,OO,4627,SFO,SLC,2025-01-02 16:43:00,2025-01-02 19:41:00,C32,C4,San Francisco,Salt Lake City
FLIGHT#0310,OO,6476,SFO,MFR,2025-01-02 16:38:00,2025-01-02 18:16:00,B26,D33,San Francisco,Medford
FLIGHT#0311,UA,486,SFO,LAS,2025-01-02 16:52:00,2025-01-02 18:24:00,D25,E8,San Francisco,Las Vegas
FLIGHT#0312,DL,214,SFO,DTW,2025-01-02 19:09:00,2025-01-03 02:20:00,C6,E35,San Francisco,Detroit
FLIGHT#0313,CY,592,SFO,PSP,2025-01-02 16:54:00,2025-01-02 18:14:00,B40,B9,San Francisco,Palm Springs
FLIGHT#0314,UA,529,SNA,SFO,2025-01-02 17:01:00,2025-01-02 18:15:00,D28,B30,Santa Ana,San Francisco
FLIGHT#0315,UA,372,MSP,SFO,2025-01-02 17:39:00,2025-01-02 19:19:00,E28,C18,Minneapolis,San Francisco
FLIGHT#0316,AA,1213,SFO,DFW,2025-01-02 18:20:00,2025-01-03 00:19:00,E39,A42,San Francisco,Dallas-Fort Worth
FLIGHT#0317,AS,341,PSP,SFO,2025-01-02 17:15:00,2025-01-02 18:37:00,D40,D15,Palm Springs,San Francisco
FLIGHT#0318,OO,5418,SBP,SFO,2025-01-02 18:11:00,2025-01-02 19:09:00,C31,D39,San Luis Obispo,San Francisco
FLIGHT#0319,CY,193,EWR,SFO,2025-01-02 17:29:00,2025-01-02 20:59:00,E20,E40,Newark,San Francisco
FLIGHT#0320,UA,1723,SFO,KOA,2025-01-02 17:26:00,2025-01-02 21:02:00,E18,D41,San Francisco,Kona
FLIGHT#0321,UA,1762,SFO,IAH,2025-01-02 17:37:00,2025-01-02 23:11:00,C40,C37,San Francisco,Houston
FLIGHT#0322,CY,720,SFO,DAL,2025-01-02 17:23:00,2025-01-02 22:47:00,D13,D23,San Francisco,Dallas
FLIGHT#0323,CY,963,SAN,SFO,2025-01-02 17:36:00,2025-01-02 19:05:00,B1,E32,San Diego,San Francisco
FLIGHT#0324,OO,5255,SFO,EUG,2025-01-02 17:27:00,2025-01-02 19:02:00,C7,B2,San Francisco,Eugene
FLIGHT#0325,CY,960,SFO,SAN,2025-01-02 17:32:00,2025-01-02 18:50:00,C3,C4,San Francisco,San Diego
FLIGHT#0326,UA,1419,MSY,SFO,2025-01-02 17:46:00,2025-01-02 20:07:00,C26,E49,New Orleans,San Francisco
FLIGHT#0327,WN,1585,SFO,LAX,2025-01-02 17:46:00,2025-01-02 19:31:00,A12,A12,San Francisco,Los Angeles
FLIGHT#0328,DL,1505,MSP,SFO,2025-01-02 17:47:00,2025-01-02 19:48:00,A33,B37,Minneapolis,San Francisco
FLIGHT#0329,US,686,PHX,SFO,2025-01-02 17:59:00,2025-01-02 19:05:00,C28,C11,Phoenix,San Francisco
FLIGHT#0330,AS,302,SEA,SFO,2025-01-02 18:00:00,2025-01-02 20:04:00,B28,E17,Seattle,San Francisco
FLIGHT#0331,UA,1552,RDU,SFO,2025-01-02 18:08:00,2025-01-02 20:51:00,B15,D48,Raleigh-durham,San Francisco
FLIGHT#0332,UA,697,LAS,SFO,2025-01-02 17:55:00,2025-01-02 19:14:00,E37,B31,Las Vegas,San Francisco
FLIGHT#0333,AA,219,LAX,SFO,2025-01-02 17:54:00,2025-01-02 19:05:00,C27,E34,Los Angeles,San Francisco
FLIGHT#0334,OO,5425,ONT,SFO,2025-01-02 18:04:00,2025-01-02 19:23:00,C46,A14,Ontario,San Francisco
FLIGHT#0335,WN,1861,SFO,PHX,2025-01-02 20:49:00,2025-01-02 23:31:00,C31,B42,San Francisco,Phoenix
FLIGHT#0336,OO,5631,SFO,BUR,2025-01-02 18:10:00,2025-01-02 19:29:00,D26,E29,San Francisco,Burbank
FLIGHT#0337,UA,720,SFO,AUS,2025-01-02 18:13:00,2025-01-02 23:41:00,E10,D1,San Francisco,Austin
FLIGHT#0338,UA,1581,SFO,OGG,2025-01-02 18:15:00,2025-01-02 22:06:00,B40,E18,San Francisco,Kahului
FLIGHT#0339,OO,5227,SFO,SAN,2025-01-02 18:26:00,2025-01-02 19:59:00,E33,E50,San Francisco,San Diego
FLIGHT#0340,CY,941,LAX,SFO,2025-01-02 18:24:00,2025-01-02 19:39:00,D4,A22,Los Angeles,San Francisco
FLIGHT#0341,AA,1459,ORD,SFO,2025-01-02 18:21:00,2025-01-02 21:11:00,D42,B25,Chicago,San Francisco
FLIGHT#0342,OO,5442,SFO,RDD,2025-01-02 18:19:00,2025-01-02 19:35:00,E21,B9,San Francisco,Redding
FLIGHT#0343,AA,69,MIA,SFO,2025-01-02 18:31:00,2025-01-02 21:33:00,B11,B15,Miami,San Francisco
FLIGHT#0344,OO,6220,DFW,SFO,2025-01-02 19:01:00,2025-01-02 20:40:00,D35,E4,Dallas-Fort Worth,San Francisco
FLIGHT#0345,UA,1670,SFO,HNL,2025-01-02 18:42:00,2025-01-02 22:52:00,A15,C18,San Francisco,Honolulu
FLIGHT#0346,DL,435,JFK,SFO,2025-01-02 18:57:00,2025-01-02 22:25:00,D35,C33,New York,San Francisco
FLIGHT#0347,UA,508,SEA,SFO,2025-01-02 19:01:00,2025-01-02 21:05:00,C16,B8,Seattle,San Francisco
FLIGHT#0348,UA,1523,BOS,SFO,2025-01-02 21:19:00,2025-01-03 01:19:00,E7,B32,Boston,San Francisco
FLIGHT#0349,CY,593,PSP,SFO,2025-01-02 19:03:00,2025-01-02 20:21:00,C13,A47,Palm Springs,San Francisco
FLIGHT#0350,UA,741,ORD,SFO,2025-01-02 19:44:00,2025-01-02 22:14:00,A16,C32,Chicago,San Francisco
FLIGHT#0351,WN,3414,SFO,SAN,2025-01-02 19:43:00,2025-01-02 21:10:00,E3,D46,San Francisco,San Diego
FLIGHT#0352,UA,404,SFO,PDX,2025-01-02 19:42:00,2025-01-02 21:24:00,B48,A30,San Francisco,Portland
FLIGHT#0353,UA,637,SFO,SNA,2025-01-02 19:32:00,2025-01-02 20:49:00,D40,A23,San Francisco,Santa Ana
FLIGHT#0354,OO,5573,SAT,SFO,2025-01-02 21:13:00,2025-01-02 22:57:00,D34,E33,San Antonio,San Francisco
FLIGHT#0355,AA,219,SFO,LAX,2025-01-02 20:03:00,2025-01-02 22:00:00,E27,D38,San Francisco,Los Angeles
FLIGHT#0356,OO,5600,BUR,SFO,2025-01-02 19:55:00,2025-01-02 21:21:00,A7,A27,Burbank,San Francisco
FLIGHT#0357,CY,947,LAX,SFO,2025-01-02 20:38:00,2025-01-02 22:06:00,D28,C15,Los Angeles,San Francisco
FLIGHT#0358,WN,138,SFO,LAS,2025-01-02 20:26:00,2025-01-02 21:49:00,B35,E31,San Francisco,Las Vegas
FLIGHT#0359,OO,5302,SFO,PSC,2025-01-02 20:29:00,2025-01-02 22:28:00,C9,A15,San Francisco,Pasco
FLIGHT#0360,CY,748,SFO,SEA,2025-01-02 20:46:00,2025-01-02 23:17:00,D7,E25,San Francisco,Seattle
FLIGHT#0361,UA,236,SFO,EWR,2025-01-02 20:41:00,2025-01-03 05:04:00,D12,B47,San Francisco,Newark
FLIGHT#0362,OO,6503,SFO,EUG,2025-01-02 20:54:00,2025-01-02 22:27:00,A50,E49,San Francisco,Eugene
FLIGHT#0363,WN,1174,LAS,SFO,2025-01-02 20:58:00,2025-01-02 22:19:00,C42,D33,Las Vegas,San Francisco
FLIGHT#0364,AS,744,SFO,SLC,2025-01-02 20:52:00,2025-01-02 23:33:00,B11,A39,San Francisco,Salt Lake City
FLIGHT#0365,WN,4992,SFO,PHX,2025-01-02 21:49:00,2025-01-03 00:32:00,D19,E15,San Francisco,Phoenix
FLIGHT#0366,UA,469,EWR,SFO,2025-01-02 23:07:00,2025-01-03 02:27:00,E3,D19,Newark,San Francisco
FLIGHT#0367,UA,1163,SFO,EWR,2025-01-02 21:44:00,2025-01-03 05:45:00,D19,C17,San Francisco,Newark
FLIGHT#0368,DL,2337,SFO,ATL,2025-01-02 21:44:00,2025-01-03 05:23:00,B4,A39,San Francisco,Atlanta
FLIGHT#0369,OO,6432,SFO,EUG,2025-01-02 22:24:00,2025-01-02 23:57:00,E11,A4,San Francisco,Eugene
FLIGHT#0370,US,898,SFO,CLT,2025-01-02 22:24:00,2025-01-03 06:16:00,A45,A24,San Francisco,Charlotte
FLIGHT#0371,UA,292,SFO,MCO,2025-01-02 22:30:00,2025-01-03 06:37:00,C7,C5,San Francisco,Orlando
FLIGHT#0372,OO,5600,SFO,MFR,2025-01-02 22:36:00,2025-01-03 00:01:00,E42,E38,San Francisco,Medford
FLIGHT#0373,UA,263,LAX,SFO,2025-01-02 22:48:00,2025-01-03 00:04:00,B22,E27,Los Angeles,San Francisco
FLIGHT#0374,OO,5414,SFO,RDM,2025-01-02 22:32:00,2025-01-03 00:12:00,A46,C18,San Francisco,Redmond-Bend
FLIGHT#0375,UA,1241,SFO,EWR,2025-01-02 23:22:00,2025-01-03 07:44:00,A28,E27,San Francisco,Newark
FLIGHT#0376,UA,522,SFO,PDX,2025-01-02 22:55:00,2025-01-03 00:42:00,A10,E39,San Francisco,Portland
FLIGHT#0377,UA,724,SFO,ORD,2025-01-02 23:05:00,2025-01-03 05:20:00,B40,D23,San Francisco,Chicago
FLIGHT#0378,UA,156,SFO,MSY,2025-01-02 23:22:00,2025-01-03 05:24:00,C40,A9,San Francisco,New Orleans
FLIGHT#0379,CY,358,SFO,BOS,2025-01-02 22:52:00,2025-01-03 07:31:00,B23,A44,San Francisco,Boston
FLIGHT#0380,DL,1370,SFO,JFK,2025-01-02 23:04:00,2025-01-03 07:36:00,E10,A49,San Francisco,New York
FLIGHT#0381,UA,1729,SFO,IAD,2025-01-02 23:10:00,2025-01-03 07:35:00,D32,E42,San Francisco,Washington
FLIGHT#0382,CY,34,SFO,JFK,2025-01-02 23:15:00,2025-01-03 07:54:00,D28,E25,San Francisco,New York
FLIGHT#0383,DL,1880,SFO,ATL,2025-01-02 00:53:00,2025-01-02 08:27:00,E2,A29,San Francisco,Atlanta
FLIGHT#0384,UA,1523,SFO,EWR,2025-01-02 23:39:00,2025-01-03 07:43:00,B9,D15,San Francisco,Newark
FLIGHT#0385,US,426,PHX,SFO,2025-01-02 00:18:00,2025-01-02 01:16:00,D47,C42,Phoenix,San Francisco

  4. Importez les données du fichier flights.csv dans la table flights :

    cbt import flights flights.csv

  5. Dans Bigtable Studio, exécutez les requêtes suivantes pour vérifier que les tables sont renseignées :

    • SELECT * FROM airports LIMIT 10;
       
    • SELECT * FROM flights LIMIT 10;

Créer un ensemble de contextes dans Bigtable Studio

Dans cette section, créez un ensemble de contextes nommé flights-assistant. Cet ensemble de contextes n'inclut aucun fichier d'ensemble de contextes importé.

  1. Dans la console Google Cloud , accédez à la page "Bigtable".

    Accéder à Bigtable

  2. Sélectionnez une instance dans la liste.

  3. Dans le menu de navigation, cliquez sur Bigtable Studio.

  4. Dans le volet Explorateur, à côté de Ensembles de contextes, cliquez sur Afficher les actions ().

  5. Cliquez sur Créer un ensemble de contextes.

  6. Dans Nom de l'ensemble de contexte, saisissez flights-assistant.

  7. Cliquez sur Créer.

Tester QueryData dans Studio

Dans cette section, utilisez l'ensemble de contexte flights-assistant en posant des questions en langage naturel pour générer une requête SQL. Comme l'ensemble de contexte ne contient aucun fichier d'ensemble de contexte importé et que les schémas Bigtable n'exposent que les noms de familles de colonnes (info) plutôt que les qualificatifs de colonnes individuels, QueryData devine les qualificatifs de colonnes et génère des requêtes non optimales.

  1. Dans le volet Explorateur, à côté de votre ensemble de contexte, cliquez sur Afficher les actions ().
  2. Cliquez sur Tester l'ensemble de contextes.
  3. Dans l'éditeur de requête, cliquez sur Générer du code SQL à l'aide de QueryData avec : flights-assistant.

  4. Saisissez la question suivante en langage naturel pour générer une requête SQL, puis cliquez sur Générer.

    Find flights from SFO to JFK.

    Examinez la requête SQL. Sans contexte défini, QueryData voit la famille de colonnes d'informations (MAP<BYTES, BYTES>) sur la table des vols, mais pas les qualificatifs de colonne departure_airport et arrival_airport. Il devine donc des noms de qualificatifs tels que origin et destination : 

      SELECT
    _key,
    info
  FROM
    flights
  WHERE
    CAST(info['origin'] AS STRING) = 'SFO'
    AND CAST(info['destination'] AS STRING) = 'JFK';

  5. Dans la fenêtre Générer du code SQL à l'aide de QueryData avec : flights-assistant, cliquez sur Modifier.

  6. Saisissez la question en langage naturel suivante pour générer une requête SQL, puis cliquez sur Mettre à jour.

    Tell me flights that can help me beat nighttime traffic if traveling from New York

    La base de données ne comprend pas le terme nighttime traffic et devine des qualificatifs de colonne qui n'existent pas dans la table flights, comme le montre la requête suivante. 

    -- Assuming 'nighttime traffic' refers to flights departing during the night,
-- defined here as between 8:00 PM (20:00) and 5:00 AM (05:00).
SELECT
  CAST(_key AS STRING) AS flight_id,
  CAST(info['flight_number'] AS STRING) AS flight_number,
  CAST(info['origin'] AS STRING) AS origin,
  CAST(info['destination'] AS STRING) AS destination,
  CAST(info['departure_time'] AS STRING) AS departure_time
FROM
  flights
WHERE
  CAST(info['origin'] AS STRING) = 'New York'
  AND (
    EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) >= 20
    OR EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) < 5
  );

  7. Dans la fenêtre Générer du code SQL à l'aide de QueryData avec : flights-assistant, cliquez sur Modifier.

  8. Saisissez la question en langage naturel suivante pour générer une requête SQL, puis cliquez sur Mettre à jour.

    Show me flights to Raleigh-Durham

    Sans contexte de recherche de valeur, QueryData devine un qualificatif de colonne et filtre sur Raleigh-Durham. Étant donné que la table flights stocke la ville sous la forme Raleigh-durham sous info['arrival_city'], la requête ne renvoie aucune ligne.

    SELECT
  _key,
  info
FROM
  flights
WHERE
  CAST(info['destination_city'] AS STRING) = 'Raleigh-Durham';

Générer un contexte pour l'ensemble de contextes

Dans cette section, vous allez préparer l'environnement et créer un fichier de contexte qui permet d'améliorer les capacités d'interrogation de l'ensemble de contexte.

Configurer votre environnement

Avant de pouvoir commencer à générer du contexte, vous devez préparer votre environnement.

Pour configurer votre environnement, procédez comme suit :

  1. Installez la CLI Antigravity. Pour en savoir plus, consultez Premiers pas avec la CLI Antigravity.
  2. Installez la Google Cloud CLI.

  3. Configurez les Identifiants par défaut de l'application (ADC). Exécutez les commandes suivantes dans votre terminal pour vous authentifier et sélectionner votre projet :

    gcloud auth application-default login

  4. Installez uv, un gestionnaire de packages et de projets Python requis pour l'agent d'ingénierie de contexte. Si Python n'est pas déjà installé sur votre système, uv peut l'installer pour vous.

  5. Installez l'agent d'ingénierie de contexte :

    agy plugin install https://github.com/GoogleCloudPlatform/db-context-enrichment/tree/main/plugin

  6. Dans votre terminal, démarrez la CLI Antigravity.

    agy

  7. Configurer une connexion à une base de données pour la génération de contexte à l'aide de MCP Toolbox. La connexion est définie dans le fichier de configuration tools.yaml.

    Pour créer le fichier de configuration tools.yaml dans votre répertoire actuel, saisissez une requête telle que : 

    Help me set up a database connection.

    Suivez les instructions fournies par la skill. Pour en savoir plus sur le fichier tools.yaml, consultez la documentation de MCP Toolbox.

  8. Pour actualiser la configuration après la création du fichier tools.yaml, exécutez la commande suivante dans la CLI Antigravity et suivez les instructions pour actualiser le serveur MCP de la boîte à outils :

    /mcp

    1. Sélectionnez db-context-engineering_toolbox.

    2. Sélectionnez Restart.

  9. Vérifiez que la boîte à outils MCP et le plug-in de l'agent d'ingénierie de contexte sont connectés et prêts à être utilisés.

    /mcp

Générer le contexte du modèle

Dans cette section, pour résoudre le problème de la section précédente où QueryData ne connaissait pas les qualificatifs de colonne flights ni le terme nighttime traffic, définissez des modèles dans le fichier d'ensemble de contexte qui mappent les questions en langage naturel aux qualificatifs de la famille de colonnes d'informations et définissez nighttime traffic comme le trafic se produisant entre 17h00 et 19h00.

Pour générer le contexte du modèle, procédez comme suit :

  1. Saisissez le prompt suivant dans la CLI Antigravity :

    Generate templates for the following two questions and SQL queries:

1. Question: "Show me flights from SFO to JFK"
SQL:
SELECT
  CAST(_key AS STRING) AS flight_id,
  CAST(info['airline'] AS STRING) AS airline,
  CAST(info['flight_number'] AS STRING) AS flight_number,
  CAST(info['departure_airport'] AS STRING) AS departure_airport,
  CAST(info['arrival_airport'] AS STRING) AS arrival_airport,
  CAST(info['departure_time'] AS STRING) AS departure_time,
  CAST(info['arrival_time'] AS STRING) AS arrival_time,
  CAST(info['departure_gate'] AS STRING) AS departure_gate,
  CAST(info['arrival_gate'] AS STRING) AS arrival_gate,
  CAST(info['departure_city'] AS STRING) AS departure_city,
  CAST(info['arrival_city'] AS STRING) AS arrival_city
FROM
  flights
WHERE
  CAST(info['departure_airport'] AS STRING) = 'SFO'
  AND CAST(info['arrival_airport'] AS STRING) = 'JFK'
ORDER BY
  departure_time;

2. Question: "Tell me flights that can help me beat nighttime traffic if traveling from New York"
SQL:
SELECT
  CAST(info['airline'] AS STRING) AS airline,
  CAST(info['flight_number'] AS STRING) AS flight_number,
  CAST(info['departure_airport'] AS STRING) AS departure_airport,
  CAST(info['departure_time'] AS STRING) AS departure_time
FROM
  flights
WHERE
  CAST(info['departure_city'] AS STRING) = 'New York'
  AND (
    EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) < 17
    OR EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) >= 19
  )
ORDER BY
  departure_time;

  2. L'agent convertit votre entrée dans un format spécifique qui affine les performances de l'ensemble de contexte sur un large éventail de requêtes utilisateur. Pour en savoir plus, consultez Présentation des ensembles de contextes.

  3. Examinez le modèle de requête généré. Vous pouvez enregistrer le modèle de requête en tant que nouveau fichier de contexte d'agent ou l'ajouter à un fichier de contexte d'agent existant.

  4. Demandez à l'agent d'enregistrer le modèle :

    Save it to context_set.json.

    L'agent crée le fichier dans le même répertoire, avec le contenu suivant :

    {
  "templates": [
    {
      "nl_query": "Show me flights from SFO to JFK",
      "sql": "SELECT CAST(_key AS STRING) AS flight_id, CAST(info['airline'] AS STRING) AS airline, CAST(info['flight_number'] AS STRING) AS flight_number, CAST(info['departure_airport'] AS STRING) AS departure_airport, CAST(info['arrival_airport'] AS STRING) AS arrival_airport, CAST(info['departure_time'] AS STRING) AS departure_time, CAST(info['arrival_time'] AS STRING) AS arrival_time, CAST(info['departure_gate'] AS STRING) AS departure_gate, CAST(info['arrival_gate'] AS STRING) AS arrival_gate, CAST(info['departure_city'] AS STRING) AS departure_city, CAST(info['arrival_city'] AS STRING) AS arrival_city FROM flights WHERE CAST(info['departure_airport'] AS STRING) = 'SFO' AND CAST(info['arrival_airport'] AS STRING) = 'JFK' ORDER BY departure_time",
      "intent": "Find flights from SFO to JFK.",
      "manifest": "Find flights from a given departure airport code to a given arrival airport code.",
      "parameterized": {
        "parameterized_sql": "SELECT CAST(_key AS STRING) AS flight_id, CAST(info['airline'] AS STRING) AS airline, CAST(info['flight_number'] AS STRING) AS flight_number, CAST(info['departure_airport'] AS STRING) AS departure_airport, CAST(info['arrival_airport'] AS STRING) AS arrival_airport, CAST(info['departure_time'] AS STRING) AS departure_time, CAST(info['arrival_time'] AS STRING) AS arrival_time, CAST(info['departure_gate'] AS STRING) AS departure_gate, CAST(info['arrival_gate'] AS STRING) AS arrival_gate, CAST(info['departure_city'] AS STRING) AS departure_city, CAST(info['arrival_city'] AS STRING) AS arrival_city FROM flights WHERE CAST(info['departure_airport'] AS STRING) = ? AND CAST(info['arrival_airport'] AS STRING) = ? ORDER BY departure_time",
        "parameterized_intent": "Find flights from ? to ?."
      }
    },
    {
      "nl_query": "Tell me flights that can help me beat nighttime traffic if traveling from New York",
      "sql": "SELECT CAST(info['airline'] AS STRING) AS airline, CAST(info['flight_number'] AS STRING) AS flight_number, CAST(info['departure_airport'] AS STRING) AS departure_airport, CAST(info['departure_time'] AS STRING) AS departure_time FROM flights WHERE CAST(info['departure_city'] AS STRING) = 'New York' AND (EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) < 17 OR EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) >= 19) ORDER BY departure_time",
      "intent": "Tell me flights that can help me beat nighttime traffic if traveling from New York",
      "manifest": "Tell me flights that can help me beat nighttime traffic if traveling from a given city",
      "parameterized": {
        "parameterized_sql": "SELECT CAST(info['airline'] AS STRING) AS airline, CAST(info['flight_number'] AS STRING) AS flight_number, CAST(info['departure_airport'] AS STRING) AS departure_airport, CAST(info['departure_time'] AS STRING) AS departure_time FROM flights WHERE CAST(info['departure_city'] AS STRING) = ? AND (EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) < 17 OR EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) >= 19) ORDER BY departure_time",
        "parameterized_intent": "Tell me flights that can help me beat nighttime traffic if traveling from ?"
      }
    }
  ]
}

Générer un contexte de recherche de valeur

Dans cette section, vous allez générer un contexte de recherche de valeur pour aider la logique de définition du contexte à mapper les expressions de valeur à des valeurs spécifiques stockées dans vos colonnes Bigtable à l'aide d'une correspondance de chaîne exacte non sensible à la casse. Par exemple, si un utilisateur recherche des vols vers "Raleigh-Durham", la recherche de valeur mappe cette requête à la valeur de base de données stockée "Raleigh-durham" (code d'aéroport RDU).

Pour générer un contexte de recherche de valeur, procédez comme suit :

  1. Saisissez le prompt suivant dans le harnais d'agent.

    Generate a case-insensitive value search on table "airports" for column "info['city']" with concept "Airport City" using match function "EXACT_MATCH_STRINGS".

  2. Examinez la définition de recherche de valeur générée. Vous pouvez enregistrer la définition de la recherche de valeurs en tant que nouveau fichier d'ensemble de contextes ou l'ajouter à un fichier d'ensemble de contextes existant.

  3. Envoyez le prompt Append the context to context_set.json à l'agent. La définition de la recherche de valeur est alors ajoutée au fichier de contexte créé dans la section précédente.

  4. Saisissez l'instance de base de données et le nom de la base de données pour lesquels le fichier d'ensemble de contexte a été généré.

    Le fichier de contexte existant est mis à jour avec la définition de recherche de valeur, avec le contenu suivant :

    {
  "templates": [
    {
      "nl_query": "Show me flights from SFO to JFK",
      "sql": "SELECT CAST(_key AS STRING) AS flight_id, CAST(info['airline'] AS STRING) AS airline, CAST(info['flight_number'] AS STRING) AS flight_number, CAST(info['departure_airport'] AS STRING) AS departure_airport, CAST(info['arrival_airport'] AS STRING) AS arrival_airport, CAST(info['departure_time'] AS STRING) AS departure_time, CAST(info['arrival_time'] AS STRING) AS arrival_time, CAST(info['departure_gate'] AS STRING) AS departure_gate, CAST(info['arrival_gate'] AS STRING) AS arrival_gate, CAST(info['departure_city'] AS STRING) AS departure_city, CAST(info['arrival_city'] AS STRING) AS arrival_city FROM flights WHERE CAST(info['departure_airport'] AS STRING) = 'SFO' AND CAST(info['arrival_airport'] AS STRING) = 'JFK' ORDER BY departure_time",
      "intent": "Find flights from SFO to JFK.",
      "manifest": "Find flights from a given departure airport code to a given arrival airport code.",
      "parameterized": {
        "parameterized_sql": "SELECT CAST(_key AS STRING) AS flight_id, CAST(info['airline'] AS STRING) AS airline, CAST(info['flight_number'] AS STRING) AS flight_number, CAST(info['departure_airport'] AS STRING) AS departure_airport, CAST(info['arrival_airport'] AS STRING) AS arrival_airport, CAST(info['departure_time'] AS STRING) AS departure_time, CAST(info['arrival_time'] AS STRING) AS arrival_time, CAST(info['departure_gate'] AS STRING) AS departure_gate, CAST(info['arrival_gate'] AS STRING) AS arrival_gate, CAST(info['departure_city'] AS STRING) AS departure_city, CAST(info['arrival_city'] AS STRING) AS arrival_city FROM flights WHERE CAST(info['departure_airport'] AS STRING) = ? AND CAST(info['arrival_airport'] AS STRING) = ? ORDER BY departure_time",
        "parameterized_intent": "Find flights from ? to ?."
      }
    },
    {
      "nl_query": "Tell me flights that can help me beat nighttime traffic if traveling from New York",
      "sql": "SELECT CAST(info['airline'] AS STRING) AS airline, CAST(info['flight_number'] AS STRING) AS flight_number, CAST(info['departure_airport'] AS STRING) AS departure_airport, CAST(info['departure_time'] AS STRING) AS departure_time FROM flights WHERE CAST(info['departure_city'] AS STRING) = 'New York' AND (EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) < 17 OR EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) >= 19) ORDER BY departure_time",
      "intent": "Tell me flights that can help me beat nighttime traffic if traveling from New York",
      "manifest": "Tell me flights that can help me beat nighttime traffic if traveling from a given city",
      "parameterized": {
        "parameterized_sql": "SELECT CAST(info['airline'] AS STRING) AS airline, CAST(info['flight_number'] AS STRING) AS flight_number, CAST(info['departure_airport'] AS STRING) AS departure_airport, CAST(info['departure_time'] AS STRING) AS departure_time FROM flights WHERE CAST(info['departure_city'] AS STRING) = ? AND (EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) < 17 OR EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) >= 19) ORDER BY departure_time",
        "parameterized_intent": "Tell me flights that can help me beat nighttime traffic if traveling from ?"
      }
    }
  ],
  "facets": [],
  "value_searches": [
    {
      "query": "SELECT CAST(T.info['city'] AS STRING) AS value, 'info:city' AS `columns`, 'Airport City' AS concept_type, 0.0 AS distance, CONCAT('iata=', CAST(T.info['iata'] AS STRING), ', name=', CAST(T.info['name'] AS STRING)) AS context FROM `airports` AS T WHERE LOWER(CAST(T.info['city'] AS STRING)) = LOWER(CAST($value AS STRING))",
      "concept_type": "Airport City",
      "description": "Exact case-insensitive match on airport city name in the airports table."
    }
  ]
}

Importer le fichier de l'ensemble de contextes dans QueryData

Dans cette section, vous allez importer le fichier d'ensemble de contexte dans QueryData afin d'améliorer les capacités de génération SQL de QueryData sur votre base de données.

Pour importer le contexte, procédez comme suit :

  1. Dans la console Google Cloud , accédez à la page "Bigtable".

    Accéder à Bigtable

  2. Sélectionnez une instance dans la liste.

  3. Dans le menu de navigation, cliquez sur Bigtable Studio.

  4. Dans le volet Explorateur, sous Ensembles de contextes et à côté de l'ensemble de contextes flights-assistant, cliquez sur l'icône Afficher les actions ().

  5. Cliquez sur Modifier l'ensemble de contextes.

  6. Cliquez sur Parcourir dans la section Importer un fichier de contexte, puis sélectionnez le fichier d'ensemble de contextes généré précédemment.

  7. Cliquez sur Enregistrer.

Générer une requête SQL à l'aide de QueryData

Dans cette section, vous allez utiliser le fichier d'ensemble de contexte que vous avez importé pour poser des questions en langage naturel. Cela vous permet de vérifier que QueryData comprend et applique correctement les définitions de termes tels que nighttime traffic et d'autres expressions associées.

Pour générer des requêtes SQL, procédez comme suit :

  1. Dans le volet Explorateur, à côté de votre ensemble de contexte, cliquez sur Afficher les actions ().
  2. Cliquez sur Tester l'ensemble de contextes.
  3. Dans l'éditeur de requête, cliquez sur Générer du code SQL à l'aide de QueryData avec : flights-assistant.

  4. Pour générer une requête SQL, saisissez la question suivante en langage naturel, puis cliquez sur Générer.

    Tell me flights that can help me beat nighttime traffic if traveling from New York

    La requête SQL générée ressemble à celle-ci :

    SELECT
  CAST(info['airline'] AS STRING) AS airline,
  CAST(info['flight_number'] AS STRING) AS flight_number,
  CAST(info['departure_airport'] AS STRING) AS departure_airport,
  CAST(info['departure_time'] AS STRING) AS departure_time
FROM
  flights
WHERE
  CAST(info['departure_city'] AS STRING) = 'New York'
  AND (
    EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) < 17
    OR EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) >= 19
  )
ORDER BY
  departure_time;

    Il s'agit de la même question que celle que vous avez ajoutée au contexte de QueryData. Notez que QueryData peut désormais interpréter correctement le terme nighttime traffic.

    Bien que le contexte provienne d'une question particulière, QueryData l'utilise pour améliorer la génération SQL pour un large éventail de questions similaires.

  5. Dans la fenêtre Générer du code SQL à l'aide de QueryData avec : flights-assistant, cliquez sur Modifier.

  6. Saisissez la question similaire suivante pour générer une requête SQL, puis cliquez sur Mettre à jour.

    What are the flights that can help me avoid evening traffic if departing from Boston

    Étant donné que la question remplace le terme nighttime traffic par un terme similaire, evening traffic, QueryData fournit une réponse cohérente à cette question en appliquant la même interprétation.

    La requête SQL générée ressemble à celle-ci :

    SELECT
  CAST(info['airline'] AS STRING) AS airline,
  CAST(info['flight_number'] AS STRING) AS flight_number,
  CAST(info['departure_airport'] AS STRING) AS departure_airport,
  CAST(info['departure_time'] AS STRING) AS departure_time
FROM
  flights
WHERE
  CAST(info['departure_city'] AS STRING) = 'Boston'
  AND (
    EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) < 17
    OR EXTRACT(HOUR FROM TIMESTAMP(CAST(info['departure_time'] AS STRING))) >= 19
  )
ORDER BY
  departure_time;

  7. Dans la fenêtre Générer du code SQL à l'aide de QueryData avec : flights-assistant, cliquez sur Modifier.

  8. Saisissez la question suivante pour générer une requête SQL, puis cliquez sur Mettre à jour.

    Show me flights to Raleigh-Durham

    La requête SQL générée ressemble à celle-ci :

    SELECT
  CAST(_key AS STRING) AS flight_id,
  CAST(info['airline'] AS STRING) AS airline,
  CAST(info['flight_number'] AS STRING) AS flight_number,
  CAST(info['departure_airport'] AS STRING) AS departure_airport,
  CAST(info['arrival_airport'] AS STRING) AS arrival_airport,
  CAST(info['departure_time'] AS STRING) AS departure_time,
  CAST(info['arrival_time'] AS STRING) AS arrival_time,
  CAST(info['departure_gate'] AS STRING) AS departure_gate,
  CAST(info['arrival_gate'] AS STRING) AS arrival_gate,
  CAST(info['departure_city'] AS STRING) AS departure_city,
  CAST(info['arrival_city'] AS STRING) AS arrival_city
FROM
  flights
WHERE
  CAST(info['arrival_city'] AS STRING) = 'Raleigh-durham'
ORDER BY
  departure_time;

    Notez que QueryData peut désormais mapper précisément Raleigh-Durham à la valeur stockée "Raleigh-durham" (RDU).

Intégrer QueryData à votre application

Dans cette section, vous allez créer un agent de données pour une application de recherche de vols. Cet agent de données fournit une interface conversationnelle pour les tables flights et airports que vous avez créées précédemment. Il explique également comment créer et intégrer cet agent de données dans votre application à l'aide d'Agent Development Kit (ADK), de l'outil MCP QueryData de Gemini Data Analytics et d'un ensemble de contextes pour améliorer la qualité des réponses.

  1. Téléchargez MCP Toolbox version 1.14.0 ou ultérieure. MCP Toolbox expose l'agent de données en tant qu'outil auquel les applications peuvent se connecter. MCP Toolbox est différent du moteur contextuel que vous avez installé précédemment, qui génère du contexte.

  2. Configurez les identifiants par défaut de l'application (ADC).

    gcloud auth application-default login

  3. Recherchez l'ID de l'ensemble de contextes que vous avez créé précédemment. Pour savoir comment trouver l'ID de l'ensemble de contextes, consultez Trouver l'ID de l'ensemble de contextes.

  4. Ajoutez ce qui suit à la configuration tools.yaml pour vous connecter à l'agent de données à l'aide de MCP Toolbox. Pour en savoir plus, consultez Source d'analyse de données Gemini et Outil QueryData d'analyse de données Gemini.

    ---
kind: source
name: gda-api-source
type: cloud-gemini-data-analytics
projectId: "PROJECT_ID"
---
kind: tool
name: cloud_gda_query_tool
type: cloud-gemini-data-analytics-query
source: gda-api-source
description: Use this tool to send natural language queries to the Gemini Data Analytics API and receive SQL, natural language answers, and explanations.
location: "REGION_ID"
context:
  datasourceReferences:
    bigtableReference:
      databaseReference:
        projectId: "PROJECT_ID"
        instanceId: "INSTANCE_ID"
      agentContextReference:
        contextSetId: "CONTEXT_SET_ID"
generationOptions:
  generateQueryResult: true
  generateNaturalLanguageAnswer: true
  generateExplanation: true
  generateDisambiguationQuestion: true

    Remplacez les éléments suivants :

    • PROJECT_ID : ID de votre projet Google Cloud .
    • REGION_ID : région de votre cluster Bigtable (par exemple, us-central1).
    • INSTANCE_ID : ID de votre instance Bigtable.
    • CONTEXT_SET_ID : ID de l'ensemble de contextes. Pour savoir comment trouver l'ID de l'ensemble de contextes, consultez Trouver l'ID de l'ensemble de contextes.

  5. Exécutez le serveur MCP Toolbox avec le fichier tools.yaml.

    ./toolbox --config PATH_TO_TOOLS_YAML

    Remplacez PATH_TO_TOOLS_YAML par le chemin d'accès au fichier tools.yaml que vous avez créé précédemment.

  6. Créez une application ADK qui appelle l'outil QueryData de Gemini Data Analytics à l'aide du SDK Python de MCP Toolbox. Pour en savoir plus sur l'utilisation du SDK Python de MCP Toolbox, consultez le guide de démarrage rapide de Toolbox. Pour en savoir plus sur l'ADK Python, consultez le guide de démarrage rapide de l'ADK.

    1. Créez un répertoire pour stocker l'application, par exemple flight-assistant-app.

    2. Accédez au répertoire flight-assistant-app.

      mkdir flight-assistant-app
cd flight-assistant-app

    3. Exécutez les commandes suivantes dans le répertoire flight-assistant-app pour créer un environnement virtuel et installer les composants requis.

      uv init
uv add google-adk toolbox-core google-genai

    4. Configurer un agent ADK

      1. Créer un agent ADK

        uv run adk create my_agent

      2. Sélectionnez le modèle gemini-2.5-flash ou un modèle plus récent.

      3. Sélectionnez Google AI, puis saisissez votre clé API Gemini. Pour savoir comment trouver votre clé API, consultez Utiliser des clés API Gemini. (Si vous n'avez pas besoin d'un essai sans frais, vous pouvez sélectionner Vertex AI et spécifier un projet Google Cloud avec la facturation activée.)

    5. Remplacez le contenu du fichier agent.py par l'exemple de code d'application Flight Data Assistant suivant.

      from typing import cast

from google.adk.agents.llm_agent import Agent
from google.adk.agents.llm_agent import ToolUnion

from toolbox_core import ToolboxSyncClient

TOOLBOX_URL = "http://127.0.0.1:5000"

INSTRUCTION = """
# ROLE
You are a friendly and factual flight data assistant. Your goal is to help users find the best flights for their needs by providing accurate information with a helpful, professional tone.
- use the Query Data Tool to answer the user's question, if the tool fails to generate a valid query, ask the user to clarify their question.

# OPERATIONAL CONSTRAINTS
- TOOL LIMITATION: You only have access to the Query Data Tool. Do not claim to have capabilities beyond what this tool provides.
- TRANSPARENCY POLICY: Maintain a seamless user experience. Never mention that you are using a tool, querying a database, or generating SQL. Frame all responses as your own direct assistance.
- SCOPE MANAGEMENT: If a user asks for something beyond your capabilities, politely state that you cannot perform that specific task. Guide the user towards what you can help with.

# COMMUNICATION STYLE
- Be concise and scannable when listing answers.
- Maintain a helpful, professional persona.

=====

# QUERY DATA TOOL

Inputs:
1. query: A natural language formulation of a database query.

Outputs: (all optional)
1. disambiguation_question: Clarification questions or comments where the tool needs the users' input.
2. generated_query: The generated query for the user query.
3. intent_explanation: An explanation for why the tool produced `generated_query`.
4. query_result: The result of executing `generated_query`.
5. natural_language_answer: The natural language answer that summarizes the `query` and `query_result`.

Usage guidance:
1. If `disambiguation_question` is produced, then solicit the needed inputs from the user and try the tool with a new `query` that has the needed clarification.
2. If `natural_language_answer` is produced, use `intent_explanation` and `generated_query` to see if you need to clarify any assumptions for the user.
3. If the tool output indicates failure or empty results, explain that clearly using the provided reasoning.
"""

client = ToolboxSyncClient(TOOLBOX_URL)

mcp_tool = client.load_tool("cloud_gda_query_tool")

root_agent = Agent(
    model="gemini-2.5-flash",  # Or newer.
    name="root_agent",
    instruction=INSTRUCTION,
    tools=cast(list[ToolUnion], [mcp_tool]),
)

  7. Pour démarrer l'application, exécutez les commandes suivantes dans le répertoire flight-assistant-app. Accédez au serveur Web ADK à l'adresse http://127.0.0.1:8000.

    uv run adk web --port 8000

  8. Saisissez du texte, par exemple hello, pour commencer à interagir avec l'agent.

    L'agent ADK répond aux questions d'ordre général et appelle les outils MCP requis.

  9. Saisissez la question suivante concernant les vols.

    How many flights depart from the west side?

    L'outil MCP est appelé pour répondre à cette question. Comme la base de données ne définit pas the west side, QueryData suppose qu'il s'agit d'une liste d'aéroports de la côte ouest et renvoie le nombre de vols et l'explication, que l'agent utilise pour clarifier l'hypothèse dans sa réponse :

    There are 253 flights departing from the west side.

Please note that "the west side" is not explicitly defined in our records, so this count assumes flights departing from major West Coast airports: Los Angeles (LAX), San Francisco (SFO), San Diego (SAN), Seattle (SEA), Portland (PDX), Las Vegas (LAS), and Phoenix (PHX). If you had a different set of airports in mind, please let me know!

  10. Saisissez une question semblable à celle du modèle de requête généré pour l'agent.

    Help me find flights from San Francisco that avoid the evening rush hour.

    En fonction du contexte QueryData ajouté précédemment, l'outil MCP comprend que evening traffic se produit entre 17h et 19h. L'outil MCP renvoie les données associées que l'agent peut utiliser pour élaborer sa réponse.

    Here are the flights departing from San Francisco that avoid the evening rush hour (defined as 5 PM to 7 PM):

* UA 1532 departing at 05:50:00
* UA 1158 departing at 05:57:00
* CY 922 departing at 06:38:00
* OO 5441 departing at 07:08:00
* UA 616 departing at 07:14:00
* AA 24 departing at 07:14:00
* B6 434 departing at 08:00:00
* AA 242 departing at 08:18:00
* UA 1739 departing at 08:22:00
* OO 6336 departing at 08:32:00
* US 1784 departing at 08:47:00
* DL 1631 departing at 09:00:00
* DL 1106 departing at 09:06:00
* OO 5427 departing at 09:06:00
* CY 352 departing at 09:25:00

  11. Saisissez une question basée sur le type de concept que vous avez ajouté dans le contexte de l'agent QueryData.

    Show me flights to Raleigh-Durham

    En fonction du contexte de recherche de valeur ajouté précédemment, l'agent comprend que "Raleigh-Durham" correspond à la ville stockée "Raleigh-durham" (code d'aéroport RDU) et renvoie les données associées que l'agent utilisera pour construire sa réponse.

    Here is the flight heading to Raleigh-Durham (RDU):

* Flight UA 1568 departs from San Francisco (SFO) at Gate A30 on 2025-01-02 at 14:55:00 and arrives at Raleigh-Durham (RDU) at Gate A26 on 2025-01-02 at 23:03:00.

Améliorer les performances de l'agent de manière itérative

L'UI Web de l'ADK vous permet d'inspecter la requête et la réponse de l'outil MCP QueryData de Gemini Data Analytics. Vous pouvez utiliser cette réponse pour observer les réponses de l'outil, telles que la requête SQL générée, l'ensemble de résultats, l'explication de l'intention, la question de désambiguïsation et la réponse en langage naturel, afin de vous aider à confirmer l'exactitude des réponses de votre agent.

Par exemple, pour le texte d'entrée How many flights depart from the west side? que vous avez saisi précédemment, cliquez sur la bulle de l'agent. Dans l'onglet Événement du panneau de navigation de gauche, développez functionResponse pour afficher la réponse suivante.

{"result": "{"generatedQuery":"SELECT COUNT(*) AS flight_count FROM flights WHERE CAST(info['departure_airport'] AS STRING) IN ('LAX', 'SFO', 'SAN', 'SEA', 'PDX', 'LAS', 'PHX')","intentExplanation":"The user is asking for the number of flights departing from the 'west side'. Since there is no explicit definition or column for 'west side' in the schema, I will assume this refers to major airports on the West Coast of the United States (e.g., LAX, SFO, SAN, SEA, PDX, LAS, PHX).","queryResult":{"columns":[{"name":"flight_count"}],"rows":[{"values":[{"value":"253"}]}],"totalRowCount":"1"},"naturalLanguageAnswer":"There are 253 flights departing from West Coast airports (LAX, SFO, SAN, SEA, PDX, LAS, PHX)."}"}

Améliorer la précision des réponses

Vous pouvez affiner en continu la précision des réponses de l'outil QueryData de Gemini Data Analytics en ajoutant du contexte. Utilisez l'agent d'ingénierie contextuelle pour générer le contexte, puis importez le fichier d'ensemble de contextes mis à jour dans l'agent de données flights-assistant existant. Pour en savoir plus, consultez Gérer les ensembles de contexte dans Bigtable Studio. La console ingère immédiatement le nouveau contexte après son importation, ce qui vous permet d'améliorer la précision de l'agent sans temps d'arrêt de l'application.

Plusieurs agents

Dans votre environnement de développement, vous pouvez effectuer des tests A/B sur plusieurs contextes d'agent de données en attribuant des noms distincts aux outils dans votre fichier tools.yaml. Par exemple, vous pouvez créer des configurations tools.yaml uniques en définissant deux outils cloud-gemini-data-analytics-query avec des noms différents, tels que cloud_gda_query_tool_v1 et cloud_gda_query_tool_v2. Cette configuration vous permet d'implémenter une logique d'application qui sélectionne de manière programmatique la version de contexte requise en choisissant le nom de l'outil correspondant.

L'exemple tools.yaml suivant montre comment configurer plusieurs agents de données pour une source de base de données :

kind: source
name: gda-api-source
type: cloud-gemini-data-analytics
projectId: PROJECT_ID
---
kind: tool
name: cloud_gda_query_tool_v1
type: cloud-gemini-data-analytics-query
source: gda-api-source
context:
  datasourceReferences:
    DB_SOURCE:
      databaseReference: ...
      agentContextReference:
        contextSetId: "V1_YOUR_CONTEXT_SET_ID"
generationOptions: ...
---
kind: tool
name: cloud_gda_query_tool_v2
type: cloud-gemini-data-analytics-query
source: gda-api-source
context:
  datasourceReferences:
    DB_SOURCE:
      databaseReference: ...
      agentContextReference:
        contextSetId: "V2_YOUR_CONTEXT_SET_ID"
generationOptions: ...

Remplacez les éléments suivants :

  • PROJECT_ID : ID de votre projet Google Cloud .
  • V1_YOUR_CONTEXT_SET_ID : ID de l'ensemble de contextes pour la version 1.
  • V2_YOUR_CONTEXT_SET_ID : ID de l'ensemble de contextes pour la version 2.
  • DB_SOURCE : configurations de la source de base de données.

Effectuer un nettoyage

Une fois que vous avez terminé ce tutoriel, supprimez les ressources que vous avez créées pour éviter des coûts récurrents. Dans les sections suivantes, nous allons voir comment supprimer ces ressources.

Supprimer l'ensemble de contexte

Avant de supprimer l'instance, supprimez l'ensemble de contextes que vous avez créé.

  1. Dans la console Google Cloud , accédez à la page "Bigtable".

    Accéder à Bigtable

  2. Sélectionnez une instance dans la liste.

  3. Dans le menu de navigation, cliquez sur Bigtable Studio.

  4. Dans le volet Explorateur, à côté de votre ensemble de contexte, cliquez sur Afficher les actions ().

  5. Cliquez sur Supprimer l'ensemble de contexte.

  6. Dans la fenêtre Supprimer l'ensemble de contextes, saisissez flights-assistant.

  7. Cliquez sur Confirmer.

Supprimer l'instance

Lorsque vous supprimez l'instance que vous avez créée dans la section Avant de commencer, vous supprimez également tous les objets que vous avez créés.

  1. Dans la console Google Cloud , accédez à la page "Bigtable".

    Accéder à Bigtable

  2. Cliquez sur le nom de l'instance que vous souhaitez supprimer, par exemple Test Instance.

  3. Cliquez sur Supprimer l'instance.

  4. Confirmez que vous souhaitez supprimer l'instance en saisissant son ID et en cliquant sur Supprimer.

Étapes suivantes