jnothman

from __future__ import print_function
import numpy as np
from sklearn.metrics import precision_recall_fscore_support as prfs, confusion_matrix
from sklearn.preprocessing import label_binarize
true = [0, 0, 0, 1, 1, 2]
preds = [('under-generate 1', [0, 0, 0, 0, 1, 2]),
         ('under-generate 2', [0, 0, 0, 1, 1, 0]),
         ('over-generate 1', [0, 1, 1, 1, 1, 2]),
         ('confuse 1 and 2', [0, 0, 0, 1, 2, 1])]

jnothman

#!/usr/bin/env python
"""
Faced with a collection of JSON blobs, this script lists what
paths (i.e. sequences of nested keys) exist in the data from
root to leaf.

For example:
  $ echo '[{"a": {"a1": 124}, "b": 111}, {"a": {"a2": 111}, "c": null}]' \
      | list-json-paths.py
will output:

jnothman

from abc import ABCMeta, abstractmethod

from .base import BaseEstimator
from .externals.six import iteritems, with_metaclass


class BaseParameterTranslator(with_metaclass(ABCMeta, BaseEstimator)):

    @property
    def fit(self):

jnothman

from __future__ import print_function, division
import numpy as np

from sklearn.base import BaseEstimator
from sklearn.cluster import MiniBatchKMeans, SpectralClustering
from sklearn.neighbors import KNeighborsClassifier
from sklearn.utils.random import sample_without_replacement
from sklearn.svm import OneClassSVM
from sklearn.linear_model import LogisticRegression
from sklearn import datasets

jnothman

#!/usr/bin/env python
"""Count or sum, while uniquing rows, without full sort of data

By using --key-fields, can also show example row that has some particular fields.

(This was much simpler when it just counted!)
"""

import sys
import argparse

jnothman

/*

On the coursera lecture index, execute this Javascript (via greasemonkey script, bookmarklet, etc) to show 'quizzes' links next to each lecture. Clicking it will open the lecture, but instead of showing the video, will:
1. Pause it
2. Show the first quiz
3. Upon clicking "skip" or "continue" on the quiz, proceed to the next
4. Repeat from 3.
5. Continue showing the video

Having watched a downloaded video, you can now easily do the in-lecture quizzes separately.

jnothman

function! SourceGitVimrc(dir)
  let gitroot = system("cd " . fnameescape(a:dir) . "; git rev-parse --show-toplevel 2>/dev/null")
  " Strip trailing newline and escape
  let gitroot = substitute(gitroot, "\\n*$","","")
  if strlen(gitroot) && filereadable(gitroot . '/.vimrc')
    execute "source " . fnameescape(gitroot) . '/.vimrc'
  endif
endfunction 

jnothman

import re, sys

num_re = re.compile(r'(?<=^\[)[0-9]+')

in_n = ''
out_n = 0

def sub_cb(match):
    global in_n, out_n

jnothman

from __future__ import print_function

from abc import ABCMeta, abstractmethod
from functools import partial

import numpy as np
from sklearn.metrics import precision_recall_fscore_support
from sklearn.base import BaseEstimator

jnothman

"""
Tool to examine the output of model selection search results from scikit-learn (assuming #1787).
Pandas might be more appropriate, but I haven't worked out how to do group_best there...

For example:
    >>> my_search = GridSearchCV(est, param_dict={'a': [...], 'b': [...], 'c': [...]})
    >>> my_search.fit(X, y)
    >>> rw = ResultsWrangler(my_search.grid_results_, my_search.fold_results_)
    >>> grouped = rw.group_best(['a', 'b'])
    >>> print(zip(grouped.parameters, grouped.scores))