SeanTater

Build a trivial FastAPI service with UV

Set up the UV project

Check out the official UV documentation but it boils down to:

• Install UV
  • on Mac and Linux it's usually just curl -LsSf https://astral.sh/uv/install.sh | sh
• Create a new project

SeanTater

-- Just as a heads up this is an experiment, not a practical script.
-- This almost always ends up in infinite loops
-- It might do a lot more compute than necessary
-- It doesn't recognize you might need more than one column to join on or it might not be as easy as =
-- But that's all beside the point.

-- What we want to show is that you can use SQL to plan how you should design your next SQL query
-- by having it use the schema (in this case fabricated) to traverse a graph.
-- Fascinating? Maybe not. But probably more practical than you expect for really really complex
-- schemata like giant businesses are prone to have.

SeanTater

#!/usr/bin/env python3
"""
    Create a video of a neural network's best estimate of a basic 2D image.
    Every frame is one epoch, and the results are saved to an MKV as VP8.
    
    In particular, this helps highlight what kinds of errors you can expect
    to see from your network based on it's representation of an easily
    understandable model.
    
    This is not by any means a good CV model and it's not intended for vision.

SeanTater

#!/usr/bin/env python3
"""
    Create a video of a neural network's best estimate of a basic 2D image.
    Every frame is one epoch, and the results are saved to an MKV as VP8.
    
    In particular, this helps highlight what kinds of errors you can expect
    to see from your network based on it's representation of an easily
    understandable model.
    
    This is not by any means a good CV model and it's not intended for vision.

SeanTater

class Result:
    def __init__(self, ok, err):
        """ Do not use this constructor. Use ok() or err(). """
        self._ok, self._err = ok, err
    
    @staticmethod
    def ok(self, ok):
        """ Create a new successful result """
        return Result(ok, None)
    

SeanTater

#!/usr/bin/env python3.7
import numpy as np
from matplotlib import pyplot
from sklearn.linear_model import SGDRegressor

L = 100
W = 20
D = 8
noise = np.random.uniform(-0.1, 0.1, L)
signal = np.sin(np.linspace(0, 10, L))

SeanTater

{-# LANGUAGE OverloadedStrings, NoImplicitPrelude #-}
import ClassyPrelude
import qualified Data.Text as Text
import qualified Data.ByteString as Bytes
import qualified Network.Download as Download
import NLP.Albemarle (Examples, Scrape, Tokens, Phrases, LSI)

main = do
  -- Lines in a (many GB) text file, autodetecting encoding
  let many_docs = sourceFile "mystuff.txt" $= Scrape.decode =$= Scrape.safeLines

SeanTater

# -*- coding: utf-8 -*-
"""
Updated on Sunday April 3 2016
@author: Marcia Price
In Python 2.7 Version

The print lines that are commented out were used for testing and 
comprehension, ie to understand how the functions worked.
"""

SeanTater

# Copyright (C) Sean Gallagher. All rights reserved. Distributed under the Apache 2 License.
# A very simple Monad (kinda)

''' A very elementary sketch of a Stream in Python.
'''
class Stream(object):
    _gen = None
    def __init__(self, gen):
        self._gen = gen

SeanTater

#!/usr/bin/env python3
import csv
import sqlite3
import argparse
import sys
import time

# Command line arguments
p = argparse.ArgumentParser(
    description="Load a CSV into an SQLite database",