vals

{
 "metadata": {
  "name": "Fitting SciLife"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {

vals

loadings = pca.components_

# I've omitted the code to create ind; a list of the indexes of the
# loadings ordered by distance from origin.

plt.scatter(*loadings, alpha=0.3, label="Loadings");
plt.scatter(*loadings[:, ind[:3]], c='r', marker='o',
            s=80, linewidths=1, facecolors="none", 
            edgecolors='r', 
            label="Contributes most\nto variance");

vals

from sklearn.decomposition.pca import PCA

pca = PCA(n_components=2)

# Note, data.shape = (6, 26440)
pca.fit(data)

Yd = dict()
Yd["400"] = pca.transform(data.ix[:,"400"])
Yd["5000"] = pca.transform(data.ix[:,"5000"])

vals

{
 "metadata": {
  "name": "Omics Data Analysis Task"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {

vals

{
 "metadata": {
  "name": "Confounding factors"
 },
 "nbformat": 3,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "heading",

vals

{
 "metadata": {
  "name": "Sandbox"
 },
 "nbformat": 3,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",

vals

###Jill Mesirov - Broad Institute

@broadinstitute

Broad has generated >150 TBPs in a year, how was this number calculated?

---

###Bas Burger - BT Global Commerce

vals

log_entry = {
 'date': '2011-03-31T00:00:00',
 'machine': 'SN167',
 'project': '0255_A81BF6ABXX',
 'size': 277025390592L
}

vals

"""Randomly picks out some title / sequence / quality triples
from a given fastq file and writes them in to a new fastq file.

This is to generate rudimentary test data which doesn't take too long to run.

Usage: make_random_test_fastq.py <fastq_file> <target file size in MB>
"""
import os
import sys
from random import random

vals

from __future__ import with_statement
import matplotlib.pyplot as plt
plt.rc('xtick', labelsize='x-small')
import yaml


def convert(yaml_file):
    """Convert the data in a yaml file generated by count_barcodes.py to a
    .dat file with space seperated sorted fractions of barcode distribution.
    """