def readInDataToPlot(self, input_fname):
"""
2009-5-20
add the column index into the column header for easy picking
2009-3-13
wrap the float conversion part into try...except to report what goes wrong
2009-3-13
"""
reader = csv.reader(open(input_fname), delimiter=figureOutDelimiter(input_fname))
self.column_header=reader.next()
for i in range(len(self.column_header)):
self.column_header[i] = '%s %s'%(i, self.column_header[i])
no_of_cols = len(self.column_header)
self.column_types = [str]*2 + [float]*(no_of_cols-2)
self.column_editable_flag_ls = [True, True] + [False]*(no_of_cols-2)
self.list_2d = []
for row in reader:
float_part = row[2:]
try:
float_part = map(float, float_part)
except:
sys.stderr.write('Except type: %s\n'%repr(sys.exc_info()))
traceback.print_exc()
new_row = row[:2]+float_part
self.list_2d.append(new_row)
self.setupColumns(self.treeview_matrix)
self.plotXY(self.ax, self.canvas, self.liststore, self.plot_title)
def run(self):
"""
2008-09-10
in case chop the whole figure into blocks, swap col_block_index and row_block_index to make row first, column 2nd
"""
from SNP import read_data
from utils import figureOutDelimiter, PassingData
delimiter = figureOutDelimiter(self.input_fname)
print delimiter
header, row_label_ls1, row_label_ls2, data_matrix = read_data(
self.input_fname, matrix_data_type=float, delimiter="\t"
)
import numpy
data_matrix = numpy.array(data_matrix)
min_value = numpy.min(data_matrix)
if self.min_value_non_negative and min_value < 0:
min_value = 0
max_value = numpy.max(data_matrix)
font = get_font(self.font_path, font_size=self.font_size)
Value2Color.special_value2color[-2] = self.super_value_color
value2color_func = lambda x: Value2Color.value2HSLcolor(x, min_value, max_value)
im_legend = drawContinousLegend(min_value, max_value, self.no_of_ticks, value2color_func, font)
fig_fname_prefix = os.path.splitext(self.fig_fname)[0]
if self.split_legend_and_matrix:
im_legend.save("%s_legend.png" % fig_fname_prefix)
no_of_rows, no_of_cols = data_matrix.shape
passParam = PassingData(
value2color_func=value2color_func,
im_legend=im_legend,
font=font,
split_legend_and_matrix=self.split_legend_and_matrix,
no_grid=self.no_grid,
)
if no_of_cols <= self.col_step_size:
self._drawMatrix(data_matrix, row_label_ls1, header[2:], self.fig_fname, passParam)
else: # split into blocks
no_of_col_blocks = no_of_cols / self.col_step_size + 1
no_of_row_blocks = no_of_rows / self.row_step_size + 1
for i in range(no_of_col_blocks):
col_start_index = i * self.col_step_size
col_end_index = (i + 1) * self.col_step_size
if col_start_index < no_of_cols:
for j in range(no_of_row_blocks):
row_start_index = j * self.row_step_size
row_end_index = (j + 1) * self.row_step_size
if row_start_index < no_of_rows:
fig_fname = "%s_%s_%s.png" % (fig_fname_prefix, j, i) # row first, column 2nd
self._drawMatrix(
data_matrix[row_start_index:row_end_index, col_start_index:col_end_index],
row_label_ls1[row_start_index:row_end_index],
header[2 + col_start_index : 2 + col_end_index],
fig_fname,
passParam,
)
def getProbeIntensityData(input_fname, data_type=numpy.float32):
"""
2010-3-18
copied from CNVNormalize.get_input()
2009-10-28
switch the default data_type to numpy.float32 to save memory on 64bit machines
2009-9-28
add argument data_type to specify data type of data_matrix.
default is numpy.float (numpy.float could be float32, float64, float128 depending on the architecture).
numpy.double is also fine.
2009-5-18
become classmethod
"""
sys.stderr.write("Getting tiling probe intensity data from %s ..."%input_fname)
import csv, subprocess
reader = csv.reader(open(input_fname), delimiter=figureOutDelimiter(input_fname))
commandline = 'wc -l %s'%input_fname
command_handler = subprocess.Popen(commandline, shell=True, stderr=subprocess.PIPE, stdout=subprocess.PIPE)
stdout_content, stderr_content = command_handler.communicate()
if stderr_content:
sys.stderr.write('stderr of %s: %s \n'%(commandline, stderr_content))
no_of_rows = int(stdout_content.split()[0])-1
header = reader.next()
no_of_cols = len(header)-3
data_matrix = numpy.zeros([no_of_rows, no_of_cols], data_type)
probe_id_ls = []
chr_pos_ls = []
i=0
for row in reader:
probe_id = row[0]
probe_id_ls.append(probe_id)
chr_pos_ls.append(tuple(row[-2:]))
for j in range(1, 1+no_of_cols):
data_matrix[i][j-1] = float(row[j])
i += 1
sys.stderr.write("Done.\n")
return data_matrix, probe_id_ls, chr_pos_ls, header
def readInRawMatrixData(self, input_fname): """ 2009-3-13 """ delimiter = figureOutDelimiter(input_fname) self.header, self.strain_acc_list, self.category_list, self.data_matrix = read_data(input_fname, delimiter=delimiter)