michaelsilverstein

#!/bin/bash -l

# USAGE: bash run_python_batchjob.qsub [PYTHON FILE] 

# Make job script
if [ -z "$1" ]
then
        echo "ERROR: MUST PROVIDE INPUT PATH"
        echo "USAGE: bash run_python_batchjob.qsub [PYTHON FILE]"
        exit

michaelsilverstein

def stackedbarplot(data, stack_order=None, palette=None, **barplot_kws):
    """
    Create a stacked barplot
    Inputs:
    | data <pd.DataFrame>: A wideform dataframe where the index is the variable to stack, the columns are different samples (x-axis), and the cells the counts (y-axis)
    | stack_order <array-like>: The order for bars to be stacked (Default: given order)
    | palette <array-like>: The colors to use for each value of `stack_order` (Default: husl)
    | barplot_kws: Arguments to pass to sns.barplot()
    
    Author: Michael Silverstein

michaelsilverstein

#!/bin/bash 

# Uses this incredible script, deinterleave_fastq.sh, and associated comments: https://gist.github.com/nathanhaigh/3521724

# Usage: deinterleave_dir.sh indir outdir

# Deinterleave entire directory of compressed .fastq.gz files from `indir` and re-compresses mates with 
# _R1.fastq.gz and _R2.fastq.gz suffixes to `ourdir`

# Download deinterleave_fastq.sh

michaelsilverstein

import matplotlib.pyplot as plt

def zebra(ax=None, orient='v', color='gray', alpha=.3, zorder=0, **kwargs):
    """
    Stripe figure - Color every other x position with `fill_between()`
    If no ax provided, use current ax
    Input:
    | ax: Axes handle
    | orient: 'v' for vertical stripes or 'h' for horizontal stripes
    | Any other argument accepted by `ax.fillbetween{x}()`

michaelsilverstein

import numpy as np
def permutation_test(a, b, n=1000):
    """
    Two-sided permutation test
    Input:
    | {a, b}: 1xm arrays of data
    | n: Number of permutations
    Output:
    | p: Two-sided pvalue: mean of |mean(a) - mean(b)| > |mean(perm(a)) - mean(perm((b)))|_i for i permutations
    

michaelsilverstein

import pandas as pd
# Data
data = [['T1', ['Round', 'Red']],
       ['T2', ['Round', 'Green']],
       ['T3', None],
       ['T4', ['Square']]]
# Make dataframe and set index
df = pd.DataFrame(data, columns=['thing', 'desc']).set_index('thing')
# "Explode" list
exploded = df.desc.apply(pd.Series)

michaelsilverstein

%%Search parameter space
[xx,yy] = meshgrid([-5:.1:9],[-5:.1:9]);

% Calculate residual for every pair of parameters
%Setup matrix where each row is a single set of parameters
X = [xx(:),yy(:)];
%Calculate residual of each pair of guesses of parameters
res = b-A*X';
r = [];
%Calculate norm^2 of residuals from each guess

michaelsilverstein

b = 

1.0e+03 *

    5.6000    2.5200    2.2400    1.9600    1.6800    1.4000    1.1200
    2.8000         0         0         0         0         0         0
    2.8000         0         0         0         0         0         0
    2.8000         0         0         0         0         0         0
    2.8000         0         0         0         0         0         0

michaelsilverstein

def pairplot(df,strain,hue=None,view='show'):
    #Pairplot http://seaborn.pydata.org/generated/seaborn.pairplot.html
    #hue: color plot by 'condition' (on). Default None (off)
    print 'Generating pairplot for %s...'%strain
    cols = list(df.columns)
    cols.remove('concentration') #No need to clutter
    if type(df.condition.iloc[0]) != str: #Apply condition name
        df.condition = df.condition.apply(lambda x: code[x])# code={0:'Wm',1:'Wc',2:'Wmc'}
    sns.pairplot(df[cols][df.strain==strain],hue=hue)
    plt.title(strain)