I used to sing a song to Emily when giving her a bath. It was a simple song with almost no tune but it made her laugh.
Scrub, scrub, scrub
Scrub the baby.
Scrub, scrub, scrub
Scrub the baby.
Scrub, scrub, scrub
Scrub the baby.
And make… a clean baby.
I don’t expect I’ll be getting a Grammy any time soon but it’s one of those silly things you do as a parent.
It was a nice but windy day so after I finished my allotted time working in the yard we went by Toys R’ Us, picked up a kite and spent a few hours down in Gully Park trying to keep it up in the air. Unfortunately it was a little too gusty and the kite a little too lightweight to stay up for long periods without divebombing into the ground. Still, it was fun and I got to answer her stream of decidedly odd questions about things like “why people who want eternal life really want eternal youth” and then trying to explain to her why I wouldn’t want either.
The you can’t see in the the pictures but the kite was a Star Wars: The Clone Wars kite with Yoda on it. Emily chose it because she knows how much I like the little Jedi Master.
She doesn’t quite get why I like Yoda more than the Sith since my wardrobe clearly marks me as one of the Dark Lords.
When Emily was a baby and an early toddler I would hold her in my hands and kick her legs while I sang the tune usually associated with the can-can dance. She used to giggle like her head was about to fall off.
A little more work on that map generator while dinner was cooking and now I have a hideous Voronoi tessellation of what was intended to be tectonic plates. I’m clearly going to have to rethink how I’m selecting the plate “centers” because I’m not getting the effect I want. Next up, Lloyd’s relaxation.
So after a couple weeks of futzing around with voronoi tessellations, locating implementations, rewriting chunks of those implementations to fit my needs, not being happy with the results, ripping it all out and starting over etc I decided I’d stick with what I Know and use a radial hex map. One cup of coffee, about[…]
It’s been a long week so I only have the basic WASD controller set up with a proper state machine. Now I can run around the featureless checkerboard plain with a few amenities like walking around a featureless checkerboard plain autorunning around a featureless checkerboard plain autowalking around a featureless checkerboard plain I’ll add jumping[…]
I spend a lot of time tinkering with various software projects, going at problems the wrong way, and generally making stupid decisions that a professional or more formally trained person would cringe to look at. Herein begins my cautionary tale as I learn painful lessons the hard way so you, dear reader, don’t have to.[…]
Coming away from the first class on the Bodhisattva Precepts at Buddha Eye Temple I have the following to roll around in my head for a while. The Bodhisattva Precepts used in the Soto Zen tradition encompass and expand on the ethics / sila encompassed in the Noble Eightfold Path. The sixteen Bodhisattva Precepts include[…]
I don’t know why I continue to use Windows. Every single god damned thing feels like pushing a giant rock up a hill. I’ll admit that the fact that I’ve been doing most of my work on Linux for the past twenty years inclines me to use the linux idiom for getting things done and[…]
HAL8999 7/100 I was sick yesterday but did spend some time looking over some “cheat sheets” that people had put together for various machine learning topics. Some were good, some were just stupid (I’m looking at you Machine Learning in Emoji). Also went through a very simply classifier based on the iris data set.
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 |
from sklearn import neighbors, datasets, preprocessing from sklearn.model_selection import train_test_split from sklearn.metrics import accuracy_score # load the data into training and test sets iris = datasets.load_iris() X, y = iris.data[:, :2], iris.target X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=13) # scale the data scaler = preprocessing.StandardScaler() scaler.fit(X_train) X_train = scaler.transform(X_train) X_test = scaler.transform(X_test) # train a k-nearest neighbors model knn = neighbors.KNeighborsClassifier(n_neighbors = 5) knn.fit(X_train, y_train) # Test the model y_pred = knn.predict(X_test) print(accuracy_score(y_test, y_pred)) |
[…]
HAL8999 6/100 Watched the “Learn how to Learn” Google talk on youtube Updated the jupyter notebooks for handson-ml from github and read through the Ch2 notebook to address the CategoricalEncoder issue from yesterday Looked at a basic transformer Part of building a data pipeline is likely to include the creation of custom transformer classes to[…]
HAL8999 – 5/100 Today while going back through the Hands On Machine Learning book Ch2 I learned that the CategoricalEncoder referenced in the section on handling categorical attributes still isn’t in scikit-learn. I checked the reqirements.txt which shows scikit-learn=0.19.1. Checking my virtualenv, I should be good.
|
1 2 3 4 |
$ grep scikit-learn handson-ml/requirements.txt scikit-learn==0.19.1 $ pip freeze |grep scikit-learn scikit-learn==0.19.2 |
Turns out that the CategoricalEncoder isn’t going to[…]
HAL8999 – [3,4]/100 Chapter 2 of Hands on ML continues Creation of test sets Stratified sampling sklearn’s StratifiedShuffleSplit Visualizing data with matplotlib Coorelation coefficients Getting a good train-test split Since you can’t train a model and just expect it to work well right out of the box it’s standard practice to split off about 20%[…]