aboutsummaryrefslogtreecommitdiff
path: root/wqflask/wqflask/comparison_bar_chart/comparison_bar_chart.py
blob: 5d74dc9d0cf12814fdf3c34b1e5d12ba569bf120 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
## Copyright (C) University of Tennessee Health Science Center, Memphis, TN.
#
# This program is free software: you can redistribute it and/or modify it
# under the terms of the GNU Affero General Public License
# as published by the Free Software Foundation, either version 3 of the
# License, or (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
# See the GNU Affero General Public License for more details.
#
# This program is available from Source Forge: at GeneNetwork Project
# (sourceforge.net/projects/genenetwork/).
#
# Contact Dr. Robert W. Williams at rwilliams@uthsc.edu
#
#
# This module is used by GeneNetwork project (www.genenetwork.org)

from __future__ import absolute_import, print_function, division

import sys
# sys.path.append(".")   Never do this in a webserver!

import string
import cPickle
import os
import time
import pp
import math
import collections
import resource


from pprint import pformat as pf

import reaper

from base.trait import create_trait
from base import data_set
from utility import webqtlUtil, helper_functions, corr_result_helpers
from db import webqtlDatabaseFunction
import utility.webqtlUtil #this is for parallel computing only.
from wqflask.correlation import correlation_functions
from utility.benchmark import Bench

from MySQLdb import escape_string as escape

from pprint import pformat as pf

from flask import Flask, g


class ComparisonBarChart(object):

    def __init__(self, start_vars):
        trait_db_list = [trait.strip() for trait in start_vars['trait_list'].split(',')]

        helper_functions.get_trait_db_obs(self, trait_db_list)

        self.all_sample_list = []
        self.traits = []
        self.insufficient_shared_samples = False
        this_group = self.trait_list[0][1].group.name #ZS: Getting initial group name before verifying all traits are in the same group in the following loop
        for trait_db in self.trait_list:
            
            if trait_db[1].group.name != this_group:
                self.insufficient_shared_samples = True
                break
            else:
                this_group = trait_db[1].group.name
            this_trait = trait_db[0]
            self.traits.append(this_trait)
            
            this_sample_data = this_trait.data

            for sample in this_sample_data:
                if sample not in self.all_sample_list:
                    self.all_sample_list.append(sample)

        if self.insufficient_shared_samples:
            pass
        else:
            self.sample_data = []
            for trait_db in self.trait_list:
                this_trait = trait_db[0]
                this_sample_data = this_trait.data

                this_trait_vals = []
                for sample in self.all_sample_list:
                    if sample in this_sample_data:
                        this_trait_vals.append(this_sample_data[sample].value)
                    else:
                        this_trait_vals.append('')
                self.sample_data.append(this_trait_vals)

            self.js_data = dict(traits = [trait.name for trait in self.traits],
                                samples = self.all_sample_list,
                                sample_data = self.sample_data,)
        
    def get_trait_db_obs(self, trait_db_list):

        self.trait_list = []
        for i, trait_db in enumerate(trait_db_list):
            if i == (len(trait_db_list) - 1):
                break
            trait_name, dataset_name = trait_db.split(":")
            #print("dataset_name:", dataset_name)
            dataset_ob = data_set.create_dataset(dataset_name)
            trait_ob = create_trait(dataset=dataset_ob,
                                   name=trait_name,
                                   cellid=None)
            self.trait_list.append((trait_ob, dataset_ob))

        #print("trait_list:", self.trait_list)