Как подключить базу данных с машинном обучением в python?

Question

Как создать бота и подключить модель pkl если у меня есть база данных [dataframe] Есть только визуализация

import bot
from tkinter import *
import csv


def enter():
    lableAnswer["text"] = bot.analiz(textCommand.get().lower())
    l = bot.list()
    lable2["text"] = bot.list_to_str(l)
    fillability = bot.fill()
    if fillability:
        lableAnswer['text'] += "\nВы согласны с результатом?"


def enter_two():
    fields = bot.list()
    with open('путь к фалу', 'a') as f:
        writer = csv.writer(f)
        writer.writerow(fields)
       # with open('titanic_train.csv', 'a', newline='') as csvFile:
       #     write = csv.writer(csvFile, delimiter=',')
       #     write.writerow([9999, 0, param[2], 'Name', param[0], param[1], 'test', 'test', 'test', 'test', 'test', 'test'])



root = Tk()
root.title('Bot')
root.geometry('600x500')

textCommand = StringVar()
questionsEntry = Entry(textvariable=textCommand)
questionsEntry.place(x=50, y=80, width=400)

lable1 = Label(text="setink")
lable1.place(x=50, y=50)

lable2 = Label(text="")
lable2.place(x=50, y=200)

lableAnswer = Label(text="")
lableAnswer.place(x=50, y=100)

button = Button(text="UP", command=enter)
button.place(x=450, y=80)

button2 = Button(text="save", command=enter_two)
button2.place(x=450, y=180)

root.mainloop()

Answer 1

#import pandas as pd # загрузка pandas
#import numpy as np
#data.columns
#data = data1[['название', 'название 1',"название","%","название 2"]]
#for i in range(1, 10):
    #data2 = data[[название', 'название 1',"название"+str(i),"%"+str(i),"знания"+str(i)]]
    #data2.columns = ['название', 'Описание',"Показатель","%","название 2"]
    #data1 = data1.append(data2)
#data1.isna().sum()
#data.dropna(inplace = True)
#data3 = data1['название 2'].str.split(';',expand=True)
#dff = pd.concat([data,data3[0]], axis = 1)
#dff.columns = ['название', 'название 1',"название","%","название 2",название 22]
#for i in range(1, 44):
    #data5 = pd.concat([data1,data3[i]], axis = 1)
    #data5.columns = [['название', 'название 1',"название","%","название 2",название 22]
#dff = dff.append(data5)
#dff["22"][0]

from sklearn.model_selection import train_test_split, cross_val_score
from sklearn.preprocessing import StandardScaler
scal = StandardScaler()
df_scal = scal.fit_transform(dtt.drop('building_age',axis=1))
y = dtt['building_age']
x = df_scal.copy()
x_train,x_test,y_train,y_test = train_test_split(x,y,test_size=0.3,random_state=13)

dt['sub_type'] = pd.factorize(dt.sub_type)[0]

Answer 2

import pandas as pd
import joblib

#dt = pd.read_csv("путь")
R_tree = joblib.load("путь")


fillability = False
sub_type = -1
listing_type = -1
tom = -1
building_age = -1
total_floor_count = -1
floor_no = -1
room_count = -1
size = -1
address = -1
heating_type = -1
pred = -1


keywords_default = [""]
keywords_sub_type = ["]
keywords_listing_type_sale = [""]
keywords_listing_type_rent = [""]
keywords_tom = [""]
keywords_building_age = [""]
keywords_total_floor_count = [""]
keywords_floor_no = [""]
keywords_room_count = [""]
keywords_size = [""]
keywords_heating_type = ["]

def fill():
    return fillability

def list():
    fields = [sub_type, listing_type, tom, building_age, total_floor_count, floor_no, room_count, size, heating_type, pred]
    return fields

def list_to_str(fields):
    str1 = ' '.join(str(e) for e in fields)
    return str1


def analiz(tex):
        global sub_type
        global listing_type
        global tom
        global building_age
        global total_floor_count
        global floor_no
        global room_count
        global size
        global heating_type
        global fillability


        text = tex
        textList = text.split()

        for i in range(len(keywords_sub_type)):
            if text.find(keywords_sub_type[i]) != -1:
                sub_type = i

        for i in range(len(textList)):
            for j in range(len(keywords_tom)):
                if textList[i].find(keywords_tom[j]) != -1:
                    tom = textList[i - 1]

        for i in range(len(keywords_listing_type_sale)):
            if text.find(keywords_listing_type_sale[i]) != -1:
                listing_type = 1

        for i in range(len(keywords_listing_type_rent)):
            if text.find(keywords_listing_type_rent[i]) != -1:
                listing_type = 2

        for i in range(len(textList)):
            for j in range(len(keywords_building_age)):
                if textList[i].find(keywords_building_age[j]) != -1:
                    building_age_l = float(textList[i - 1])
                    if building_age_l == 0:
                        building_age = 0
                    elif building_age_l == 1:
                        building_age = 1
                   

        for i in range(len(textList)):
            for j in range(len(keywords_total_floor_count)):
                if textList[i].find(keywords_total_floor_count[j]) != -1:
                    total_floor_count_l = float(textList[i - 1])
                    if total_floor_count_l == 1:
                        total_floor_count = 0
                    elif total_floor_count_l == 10:
                        total_floor_count = 1
           

        for i in range(len(textList)):
            for j in range(len(keywords_floor_no)):
                if textList[i].find(keywords_floor_no[j]) != -1:
                    floor_no_l = float(textList[i - 1])
                    if floor_no_l >= 20:
                        floor_no = 12
                    elif floor_no_l >= 10:
                        floor_no = floor_no_l - 9
                    

        for i in range(len(textList)):
            for j in range(len(keywords_room_count)):
                if textList[i].find(keywords_room_count[j]) != -1:
                    room_count = textList[i - 1]

        for i in range(len(textList)):
            for j in range(len(keywords_size)):
                if textList[i].find(keywords_size[j]) != -1:
                    size = textList[i - 1]

        for i in range(len(keywords_heating_type)):
            if text.find(keywords_heating_type[i]) != -1:
                heating_type = i

        if sub_type == -1:
            fillability = False
            return "Укажите пожалуйста тип "

        else:
            fillability = True

        if listing_type == -1:
            fillability = False
            return

        else:
            fillability = True

        if tom == -1:
            fillability = False
            return

        else:
            fillability = True

        if building_age == -1:
            fillability = False
            return

        else:
            fillability = True

        if total_floor_count == -1:
            fillability = False
            return 

        else:
            fillability = True

        if floor_no == -1:
            fillability = False
            return

        else:
            fillability = True

        if room_count == -1:
            fillability = False
            return 

        else:
            fillability = True

        if size == -1:
            fillability = False
            return 

        else:
            fillability = True

        if heating_type == -1:
            fillability = False
            return

        else:
            fillability = True


        if fillability:
            data = {"sub_type": [sub_type], 'start_date': [18], 'end_date':[19],'listing_type': [listing_type], 'tom': [tom],
                    'building_age': [building_age], 'total_floor_count': [total_floor_count], 'floor_no': [floor_no],
                    'room_count': [room_count], 'size': [size],'address':[123], 'heating_type': [heating_type]}
            df_new = pd.DataFrame(data)
            pred = R_tree.predict(df_new)
            return " " + str(pred)

БЛОГ НА HUSL

Как подключить базу данных с машинном обучением в python?

Ответы (2 шт):