danielparton

Linear Regression - lecture series II-IV

Model representation

Notation (used throughout course)

  m = number of training samples x = “input” variable / feature y = “output” variable / “target” variable (x,y) = one training example

danielparton

# for each template:
    # from PDB entry:
        # print SS bonds within template span

import os
import gzip
import ensembler
from ensembler.uniprot import get_uniprot_xml
from ensembler.initproject import extract_template_pdbchains_from_uniprot_xml, parse_sifts_xml
from ensembler.utils import set_loglevel

danielparton

import ast
import re
import operator as op
import simtk.unit

unit_membernames = [name for name in simtk.unit.__dict__]

quantity_as_number_space_unit_regex = re.compile(
    '([0-9.]+) ?({0})'.format('|'.join(unit_membernames))
)   # e.g. "2 picoseconds"

danielparton

def retrieve_uniprot(search_string, maxreadlength=100000000):
    '''
    Searches the UniProt database given a search string, and retrieves an XML
    file, which is returned as a string.
    maxreadlength is the maximum size in bytes which will be read from the website
    (default 100MB)
    Example search string: 'domain:"Protein kinase" AND reviewed:yes'

    The function also removes the xmlns attribute from <uniprot> tag, as this
    makes xpath searching annoying

danielparton

#!/usr/bin/env python
"""

Construct a configfile for MPICH2 mpirun from Torque/Moab $PBS_GPUFILE contents.

Usage:

python build-mpirun-configfile.py executable [args...]
mpirun -configfile configfile

danielparton

import socket
import mpi4py.MPI
import gzip
import simtk.openmm as openmm
import simtk.openmm.app as app
import simtk.unit as unit

comm = mpi4py.MPI.COMM_WORLD
rank = comm.rank
size = comm.size

danielparton

tokenized_string = "2 + 3 * 4".split()
 
# parsing arithmetic expressions with + and *
# A parser returns (parsed_thing, new_index) on success.
 
def fail(): raise Exception("failed to parse")
 
def is_num(x):
    try: int(x)
    except ValueError: return False