mkowoods

-- https://stackoverflow.com/a/28324562
-- https://sqlperformance.com/2013/03/io-subsystem/chunk-deletes
DECLARE @Deleted_Rows INT;
SET @Deleted_Rows = 1;


WHILE (@Deleted_Rows > 0)
  BEGIN

   BEGIN TRANSACTION

mkowoods

Source Data Set: https://console.cloud.google.com/bigquery?p=data-to-insights&d=ecommerce&t=web_analytics&page=table


## TRAIN

CREATE OR REPLACE MODEL `ecommerce.classification_model_2`
OPTIONS
  (model_type='logistic_reg', labels = ['will_buy_on_return_visit']) AS

WITH all_visitor_stats AS (

mkowoods

#snippet for getting data to and from google drive while using google colab

!pip install -q tqdm
import tqdm
from google.colab import auth
from googleapiclient.discovery import build
import io
from googleapiclient.http import MediaIoBaseDownload
from googleapiclient.http import MediaFileUpload

mkowoods

--last mod date for all tables
SELECT 
	t.name,
	t.modify_date,
	i.rowcnt
FROM sys.tables t
left outer join sysindexes i
ON t.object_id = i.id 
WHERE i.indid < 2

mkowoods

""""
uses policy/value iteration to solve the task
there's a 2 layer neural network that's responsible for estimating the future value of a state 
and a linear weight model (wrapped in a softmax) to handle selecting the policy(acitons)

the policy model incorporates the value function, by adjusting the loss score by the difference between the observed reward and the expected reward
actions that perform much better than expected will have a lower "penalty" and thus have their weights less affected than actions where the 
observed performance is much worse.

The other interesting component of this model is the exploration function, which ensures that the agent explores at a rate that's proportional 

mkowoods

import numpy as np
import gym



"""
Solved Frozen Lake: after 1100 steps.
Solved Taxi-v1: after 500 steps
Implementation of SARSA...

mkowoods

import numpy as np
import gym



"""
Implementation of SARSA...
Using a relatively small learning rate, which causes convergence to take longer
Built as an object to allow for easy extensibility to other environments for quick benchmarking

mkowoods

import os
os.chdir('/Users/mwoods/Desktop')



# -----------------
# User Instructions
# 

mkowoods

# Finite State Machine
# An Example of Finite State Machine Using Classes
# Responsed to: https://www.reddit.com/r/dailyprogrammer/comments/4cb7eh
class FiniteStateMachine:

    def __init__(self):
        self.states = set([])
        self.transitions = {}
        self.current_state = None

mkowoods

import random
import os


random.seed(42)

F_PATH = "data.csv"
OUTPUT_PATH_TEMPLATE = "tmp_%d.csv"
CHUNK_SIZE = 5