Prediction, Tables samples (#12)

- Prediction, Tables samples (32429f83624047758c1fd8c13691150eb9aaefc3)


Co-authored-by: Francesco Giannuzzo <[email protected]>

app.py

@@ -1,6 +1,23 @@
+import os
+import sys
+import time
+import re
+import csv
 import gradio as gr
 import pandas as pd
-import os
+import numpy as np
+import plotly.express as px
+import plotly.graph_objects as go
+import plotly.colors as pc
+from qatch.connectors.sqlite_connector import SqliteConnector
+from qatch.generate_dataset.orchestrator_generator import OrchestratorGenerator
+from qatch.evaluate_dataset.orchestrator_evaluator import OrchestratorEvaluator
+from prediction import ModelPrediction
+import utils_get_db_tables_info
+import utilities as us
+# @spaces.GPU
+# def model_prediction():
+#   pass
 # # https://discuss.huggingface.co/t/issues-with-sadtalker-zerogpu-spaces-inquiry-about-community-grant/110625/10
 # if os.environ.get("SPACES_ZERO_GPU") is not None:
 #     import spaces
@@ -11,23 +28,6 @@ import os
 #             def wrapper(*args, **kwargs):
 #                 return func(*args, **kwargs)
 #             return wrapper
-import sys
-from qatch.connectors.sqlite_connector import SqliteConnector
-from qatch.generate_dataset.orchestrator_generator import OrchestratorGenerator
-from qatch.evaluate_dataset.orchestrator_evaluator import OrchestratorEvaluator
-from prediction import ModelPrediction
-import utils_get_db_tables_info
-import utilities as us
-import time
-import plotly.express as px
-import plotly.graph_objects as go
-import plotly.colors as pc
-import re
-import csv
-import numpy as np
-# @spaces.GPU
-# def model_prediction():
-#   pass
 pnp_path = os.path.join(".", "evaluation_p_np_metrics.csv")
 
 js_func = """
@@ -50,8 +50,8 @@ df_default = pd.DataFrame({
     'Age': [25, 30, 35],
     'City': ['New York', 'Los Angeles', 'Chicago']
 })
-
-models_path = "./models.csv"
+models_path = os.path.join(".", "models.csv")
+#models_path = "./models.csv"
 
 # Variabile globale per tenere traccia dei dati correnti
 df_current = df_default.copy()
@@ -176,7 +176,7 @@ with gr.Blocks(theme='shivi/calm_seafoam', css_paths='style.css', js=js_func) as
     with gr.Row():
         with gr.Column(scale=1):
             gr.Image(
-                value="./qatch_logo.png",
+                value=os.path.join(".", "qatch_logo.png"),
                 show_label=False,
                 container=False,
                 interactive=False,
@@ -652,10 +652,12 @@ with gr.Blocks(theme='shivi/calm_seafoam', css_paths='style.css', js=js_func) as
                         samples = us.generate_some_samples(input_data['data']['db'], row["tbl_name"])
                         prompt_to_send = us.prepare_prompt(input_data["prompt"], question, schema_text, samples)
                         #PREDICTION SQL
-                        #response = predictor.make_prediction(question=question, db_schema=schema_text, model_name=model, prompt=f"{prompt_to_send}")
-                        prediction = row["query"]#"SQL"#response[response_parsed]
-                        price = 0.0#response[cost]
-                        answer = "Answer"#response[response]
+                        if prompt_to_send == prompt_default:
+                            prompt_to_send = None
+                        response = predictor.make_prediction(question=question, db_schema=schema_text, model_name=model, prompt=f"{prompt_to_send}")
+                        prediction = response['response_parsed']
+                        price = response['cost']
+                        answer = response['response']
 
                         end_time = time.time()
                         display_prediction = f"""<div class='loading' style='font-size: 1.7rem; font-family: 'Inter', sans-serif;'>>Predicted SQL:</div> 

tables_dict.csv

@@ -0,0 +1,300 @@
+Table: accountFraud
+hasothercards,housingstatus,dateofbirthdistinctemails4w,income,paymenttype,employmentstatus,creditriskscore,sessionlengthminutes,deviceos,emailisfree
+False,BD,3,0.5,AB,CC,230,5.07759998143027,linux,Paid
+False,BC,15,0.1,AD,CB,-40,4.0223824396911,other,Paid
+True,BC,7,0.9,AC,CA,215,3.749706225590873,windows,Paid
+False,BC,2,0.4,AC,CD,51,4.886676763177824,other,Paid
+False,BC,6,0.4,AB,CA,108,1.7508864007811145,other,Free
+
+Table: shop
+Shop_ID,Name,Location,District,Number_products,Manager_name
+1,FC Haka,Valkeakoski,Tehtaan kenttä,3516,Olli Huttunen
+2,HJK,Helsinki,Finnair Stadium,10770,Antti Muurinen
+3,FC Honka,Espoo,Tapiolan Urheilupuisto,6000,Mika Lehkosuo
+4,FC Inter,Turku,Veritas Stadion,10000,Job Dragtsma
+5,FF Jaro,Jakobstad,Jakobstads Centralplan,5000,Mika Laurikainen
+
+Table: airlines
+uid,Airline,Abbreviation,Country
+1,United Airlines,UAL,USA
+2,US Airways,USAir,USA
+3,Delta Airlines,Delta,USA
+4,Southwest Airlines,Southwest,USA
+5,American Airlines,American,USA
+
+Table: fitnessTrackers
+brandname,devicetype,modelname,color,sellingprice,originalprice,display,rating,strapmaterial,averagebatterylife
+huami,Smartwatch,Amazfit Bip S,White ,4999.0,5999.0,AMOLED Display,4.0,Thermoplastic polyurethane,15
+GARMIN ,Smartwatch,Approach S62,"White, Black",46990.0,51990.0,OLED Display,4.1,Silicone,12
+SAMSUNG ,Smartwatch,Galaxy Classic 4,Black,32959.0,37999.0,AMOLED Display,4.6,Elastomer,14
+SAMSUNG ,Smartwatch,Galaxy Watch 4 LTE,Black,31999.0,34999.0,AMOLED Display,4.6,Elastomer,14
+APPLE,Smartwatch,Series 6 GPS + Cellular 40 mm Red Aluminium Case,Red,45690.0,49900.0,OLED Retina Display,4.5,Aluminium,1
+
+Table: Ref_Template_Types
+Template_Type_Code,Template_Type_Description
+PPT,Presentation
+CV,CV
+AD,Advertisement
+PP,Paper
+BK,Book
+
+Table: stadium
+Stadium_ID,Location,Name,Capacity,Highest,Lowest,Average
+1,Raith Rovers,Stark's Park,10104,4812,1294,2106
+2,Ayr United,Somerset Park,11998,2363,1057,1477
+3,East Fife,Bayview Stadium,2000,1980,533,864
+4,Queen's Park,Hampden Park,52500,1763,466,730
+5,Stirling Albion,Forthbank Stadium,3808,1125,404,642
+
+Table: Student
+StuID,LName,Fname,Age,Sex,Major,Advisor,city_code
+1001,Smith,Linda,18,F,600,1121,BAL
+1002,Kim,Tracy,19,F,600,7712,HKG
+1003,Jones,Shiela,21,F,600,7792,WAS
+1004,Kumar,Dinesh,20,M,600,8423,CHI
+1005,Gompers,Paul,26,M,600,1121,YYZ
+
+Table: Templates
+Template_ID,Version_Number,Template_Type_Code,Date_Effective_From,Date_Effective_To,Template_Details
+0,5,PP,2005-11-12 07:09:48,2008-01-05 14:19:28,
+1,9,PP,2010-09-24 01:15:11,1999-07-08 03:31:04,
+4,4,BK,2002-03-02 14:39:49,2001-04-18 09:29:52,
+6,2,PPT,1975-05-20 22:51:19,1992-05-02 20:06:11,
+7,8,PPT,1993-10-07 02:33:04,1975-07-16 04:52:10,
+
+Table: employee
+Employee_ID,Name,Age,City
+1,George Chuter,23,Bristol
+2,Lee Mears,29,Bath
+3,Mark Regan,43,Bristol
+4,Jason Hobson,30,Bristol
+5,Tim Payne,29,Wasps
+
+Table: flights
+Airline,FlightNo,SourceAirport,DestAirport
+1,28, APG, ASY
+1,29, ASY, APG
+1,44, CVO, ACV
+1,45, ACV, CVO
+1,54, AHD, AHT
+
+Table: model_list
+ModelId,Maker,Model
+1,1,amc
+2,2,audi
+3,3,bmw
+4,4,buick
+5,4,cadillac
+
+Table: heartAttack
+age,sex,chestpaintype,restingbloodpressure,cholestoralinmg,fastingbloodsugar,restingelectrocardiographicrresults,numberofmajorvvessels,thall,output
+62,male,typicalAngina,140,268,false,normal,2,2,noHeartAttack
+64,female,asymptomatic,110,211,false,normal,0,2,heartAttack
+68,female,nonAnginalPain,180,274,true,normal,0,3,noHeartAttack
+50,female,typicalAngina,144,200,false,normal,0,3,noHeartAttack
+76,male,nonAnginalPain,140,197,false,leftVentricularHypertrophy,0,2,heartAttack
+
+Table: Paragraphs
+Paragraph_ID,Document_ID,Paragraph_Text,Other_Details
+7,2394,Korea,
+9,3,Somalia,
+65,50123,Palestinian Territory,
+241,651512,Jersey,
+3708,33930,UK,
+
+Table: concert
+concert_ID,concert_Name,Theme,Stadium_ID,Year
+1,Auditions,Free choice,1,2014
+2,Super bootcamp,Free choice 2,2,2014
+3,Home Visits,Bleeding Love,2,2015
+4,Week 1,Wide Awake,10,2014
+5,Week 1,Happy Tonight,9,2015
+
+Table: car_makers
+Id,Maker,FullName,Country
+1,amc,American Motor Company,1
+2,volkswagen,Volkswagen,2
+3,bmw,BMW,2
+4,gm,General Motors,1
+5,ford,Ford Motor Company,1
+
+Table: evaluation
+Employee_ID,Year_awarded,Bonus
+1,2011,3000.0
+2,2015,3200.0
+1,2016,2900.0
+4,2017,3200.0
+7,2018,3200.0
+
+Table: singer_in_concert
+concert_ID,Singer_ID
+1,2
+1,3
+1,5
+2,3
+2,6
+
+Table: Documents
+Document_ID,Template_ID,Document_Name,Document_Description,Other_Details
+0,7,Introduction of OS,n,
+1,25,Understanding DB,y,
+3,6,Summer Show,u,
+76,20,Robbin CV,y,
+80,14,Welcome to NY,h,
+
+Table: visit
+Museum_ID,visitor_ID,Num_of_Ticket,Total_spent
+1,5,20,320.14
+2,5,4,89.98
+4,3,10,320.44
+2,3,24,209.98
+4,6,3,20.44
+
+Table: museum
+Museum_ID,Name,Num_of_Staff,Open_Year
+1,Plaza Museum,62,2000
+2,Capital Plaza Museum,25,2012
+3,Jefferson Development Museum,18,2010
+4,Willow Grande Museum,17,2011
+5,RiverPark Museum,16,2008
+
+Table: singer
+Singer_ID,Name,Country,Song_Name,Song_release_year,Age,Is_male
+1,Joe Sharp,Netherlands,You,1992,52,F
+2,Timbaland,United States,Dangerous,2008,32,T
+3,Justin Brown,France,Hey Oh,2013,29,T
+4,Rose White,France,Sun,2003,41,F
+5,John Nizinik,France,Gentleman,2014,43,T
+
+Table: countries
+CountryId,CountryName,Continent
+1,usa,1
+2,germany,2
+3,france,2
+4,japan,3
+5,italy,2
+
+Table: continents
+ContId,Continent
+1,america
+2,europe
+3,asia
+4,africa
+5,australia
+
+Table: adultCensus
+workclass,education,maritalstatus,occupation,relationship,race,sex,hoursperweek,nativecountry,income
+Private,Doctorate,Married-civ-spouse,Prof-specialty,Husband,White,Male,60,United-States,>50K
+Local-gov,Some-college,Divorced,Adm-clerical,Not-in-family,Black,Female,40,United-States,<=50K
+Private,10th,Married-civ-spouse,Machine-op-inspct,Husband,White,Male,40,United-States,<=50K
+Private,Some-college,Married-civ-spouse,Exec-managerial,Husband,White,Male,44,United-States,>50K
+?,HS-grad,Never-married,?,Not-in-family,White,Female,40,United-States,<=50K
+
+Table: latePayment
+customerid,paperlessdate,invoicenumber,invoicedate,duedate,invoiceamount,disputed,paperlessbill,daystosettle,dayslate
+7758-WKLVM,2/6/2012,5510823569,4/11/2012,5/11/2012,30.06,No,Electronic,36,6
+2820-XGXSB,1/26/2012,6528247418,1/4/2013,2/3/2013,84.86,No,Electronic,4,0
+2621-XCLEH,9/24/2012,9465847338,6/18/2013,7/18/2013,37.49,No,Electronic,29,0
+5196-TWQXF,6/20/2012,7092718520,12/2/2012,1/1/2013,50.17,No,Electronic,22,0
+9725-EZTEJ,3/17/2012,685917930,10/13/2012,11/12/2012,124.38,Yes,Electronic,28,0
+
+Table: course_arrange
+Course_ID,Teacher_ID,Grade
+2,5,1
+2,3,3
+3,2,5
+4,6,7
+5,6,1
+
+Table: Has_Pet
+StuID,PetID
+1001,2001
+1002,2002
+1002,2003
+
+Table: breastCancer
+patientidentifier,age,menopausalstatus,tumorsize,tumorgrade,numberpositivelymphnodes,progesteronereceptor,estrogenreceptor,hormonaltherapy,status
+1522,32,premenopausal,25,2,2,36,10,no,"aliveWithoutRecurrence,"
+1329,61,postmenopausal,30,2,1,24,38,yes,"aliveWithoutRecurrence,"
+1141,62,postmenopausal,33,1,5,239,76,no,recurrenceOrDeath
+1488,66,postmenopausal,42,3,11,412,339,yes,recurrenceOrDeath
+1294,38,premenopausal,23,3,3,14,6,no,"aliveWithoutRecurrence,"
+
+Table: hiring
+Shop_ID,Employee_ID,Start_from,Is_full_time
+1,1,2009,T
+1,2,2003,T
+8,3,2011,F
+4,4,2012,T
+5,5,2013,T
+
+Table: cars_data
+Id,MPG,Cylinders,Edispl,Horsepower,Weight,Accelerate,Year
+1,18,8,307.0,130,3504,12.0,1970
+2,15,8,350.0,165,3693,11.5,1970
+3,18,8,318.0,150,3436,11.0,1970
+4,16,8,304.0,150,3433,12.0,1970
+5,17,8,302.0,140,3449,10.5,1970
+
+Table: car_names
+MakeId,Model,Make
+1,chevrolet,chevrolet chevelle malibu
+2,buick,buick skylark 320
+3,plymouth,plymouth satellite
+4,amc,amc rebel sst
+5,ford,ford torino
+
+Table: visitor
+ID,Name,Level_of_membership,Age
+1,Gonzalo Higuaín ,8,35
+2,Guti Midfielder,5,28
+3,Arjen Robben,1,27
+4,Raúl Brown,2,56
+5,Fernando Gago,6,36
+
+Table: salesTransactions
+transactionno,date,productno,productname,price,quantity,customerno,country
+549047,4/6/2019,21544,Skulls-Water-Transfer-Tattoos,11.12,2,17783.0,United-Kingdom
+566959,9/15/2019,22113,Grey-Heart-Hot-Water-Bottle,14.61,2,17530.0,United-Kingdom
+536526,12/1/2018,22173,Metal-4-Hook-Hanger-French-Chateau,13.27,16,14001.0,United-Kingdom
+580355,12/2/2019,47599A,Pink-Party-Bags,6.19,12,13953.0,United-Kingdom
+579557,11/30/2019,21671,Red-Spot-Ceramic-Drawer-Knob,6.19,1,12557.0,United-Kingdom
+
+Table: airports
+City,AirportCode,AirportName,Country,CountryAbbrev
+Aberdeen ,APG,Phillips AAF ,United States ,US 
+Aberdeen ,ABR,Municipal ,United States ,US
+Abilene ,DYS,Dyess AFB ,United States ,US
+Abilene ,ABI,Municipal ,United States ,US
+Abingdon ,VJI,Virginia Highlands ,United States ,US
+
+Table: mushrooms
+class,capshape,capsurface,capcolor,bruises,odor,gillattachment,gillspacing,gillsize,gillcolor
+poisonous,flat,smooth,white,bruises,pungent,free,close,narrow,brown
+edible,convex,fibrous,gray,no,none,free,crowded,broad,black
+edible,bell,smooth,white,bruises,almond,free,close,broad,gray
+edible,flat,scaly,brown,bruises,none,free,close,broad,pink
+poisonous,convex,fibrous,yellow,no,foul,free,close,broad,gray
+
+Table: Pets
+PetID,PetType,pet_age,weight
+2001,cat,3,12.0
+2002,dog,2,13.4
+2003,dog,1,9.3
+
+Table: teacher
+Teacher_ID,Name,Age,Hometown
+1,Joseph Huts,32,Blackrod Urban District
+2,Gustaaf Deloor,29,Bolton County Borough
+3,Vicente Carretero,26,Farnworth Municipal Borough
+4,John Deloor,33,Horwich Urban District
+5,Kearsley Brown,45,Kearsley Urban District
+
+Table: course
+Course_ID,Staring_Date,Course
+1,5 May,Language Arts
+2,6 May,Math
+3,7 May,Science
+4,9 May,History
+5,10 May,Bible
+

tables_dict.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a05f1d65a8f0d5ea2bb1aa1efd76cddb2685ca2da4a3aac54838d5be1205cbe6
+size 22799

utilities.py

@@ -1,6 +1,7 @@
 import csv
 import re
 import pandas as pd
+import pickle
 import sqlite3
 import gradio as gr
 import os
@@ -107,6 +108,10 @@ def generate_some_samples(connector, tbl_name):
         samples.append(f"Error: {e}")
     return samples
 
+def load_tables_dict(file_path):
+    with open(file_path, 'rb') as f:
+        return pickle.load(f)
+
 def extract_tables_dict(pnp_path):
     tables_dict = {}
     with open(pnp_path, mode='r', encoding='utf-8') as file:
@@ -126,5 +131,5 @@ def extract_tables_dict(pnp_path):
                     tables_dict[tbl_name] = df
                 except Exception as e:
                     tables_dict[tbl_name] = pd.DataFrame({"Error": [str(e)]})  # DataFrame con messaggio di errore
-    
-    return tables_dict
+    return load_tables_dict('tables_dict.csv')
+    return tables_dict