'''

***************************************************

v''' + version_nb + '''

author: Jean Helie (jean.helie@bioch.ox.ac.uk)

git: https://github.com/jhelie/xvg_average_residues

***************************************************

[ DESCRIPTION ]

This script calculate the average of residues ratios from residues densities contained

in several xvg files.

It calculates the avg and (unbiased) std dev and can deal with NaN.

The ratios are calculated for each supplied file, then averaged proportional to the weight

of that file.

NB:

the script may give out a warning 'return np.mean(x,axis)/factor', it's ok. it's just

scipy warning us that there were only nans on a row, the result will be a nan as we

expect (see this thread: https://github.com/scipy/scipy/issues/2898).

[ USAGE ]

Option Default Description

-----------------------------------------------------

-f : xvg file(s)

-o residues_avg : name of outptut file

--comments @,# : lines starting with these characters will be considered as comment

Other options

-----------------------------------------------------

--version : show version number and exit

-h, --help : show this menu and exit

global nb_rows

global nb_cols

global weights

global distance

global data_ratios

nb_rows = 0

nb_cols = 0

weights = np.ones(len(args.xvgfilenames))

for f_index in range(0,len(args.xvgfilenames)):

#display progress

progress = '\r -reading file ' + str(f_index+1) + '/' + str(len(args.xvgfilenames)) + ' '

sys.stdout.flush()

sys.stdout.write(progress)

#get file content

filename = args.xvgfilenames[f_index]

with open(filename) as f:

lines = f.readlines()

#determine legends and nb of lines to skip

tmp_nb_rows_to_skip = 0

for l_index in range(0,len(lines)):

line = lines[l_index]

if line[0] in args.comments:

tmp_nb_rows_to_skip += 1

if "weight" in line:

if "-> weight = " in line:

weights[f_index] = float(line.split("-> weight = ")[1])

if weights[f_index] < 0:

print "\nError: the weight in file " + str(filename) + " should be a positive number."

print " -> " + str(line)

sys.exit(1)

else:

print "\nWarning: keyword 'weight' found in the comments of file " + str(filename) + ", but weight not read in as the format '-> weight = ' wasn't found."

#get data

tmp_data = np.loadtxt(filename, skiprows = tmp_nb_rows_to_skip)

#check that each file has the same number of data rows

if f_index == 0:

nb_rows = np.shape(tmp_data)[0]

distance = np.zeros((nb_rows, 1)) #distance from cluster

data_residues_basic = np.zeros((nb_rows, len(args.xvgfilenames)))

data_residues_hphob = np.zeros((nb_rows, len(args.xvgfilenames)))

data_residues_total = np.zeros((nb_rows, len(args.xvgfilenames)))

else:

if np.shape(tmp_data)[0] != nb_rows:

print "Error: file " + str(filename) + " has " + str(np.shape(tmp_data)[0]) + " data rows, whereas file " + str(args.xvgfilenames[0]) + " has " + str(nb_rows) + " data rows."

sys.exit(1)

#check that each file has the same number of columns

if f_index == 0:

nb_cols = np.shape(tmp_data)[1]

else:

if np.shape(tmp_data)[1] != nb_cols:

print "Error: file " + str(filename) + " has " + str(np.shape(tmp_data)[1]) + " data columns, whereas file " + str(args.xvgfilenames[0]) + " has " + str(nb_cols) + " data columns."

sys.exit(1)

#check that each file has the same first column

if f_index == 0:

distance[:,0] = tmp_data[:,0]

else:

if not np.array_equal(tmp_data[:,0],distance[:,0]):

print "\nError: the first column of file " + str(filename) + " is different than that of " + str(args.xvgfilenames[0]) + "."

sys.exit(1)

#store data

data_residues_basic[:,f_index] = tmp_data[:,1]

data_residues_hphob[:,f_index] = tmp_data[:,3]

data_residues_total[:,f_index] = tmp_data[:,5]

#remove non meaningful zeros (i.e. not sampled)

data_residues_total[data_residues_total == 0] = np.nan

data_residues_basic[np.isnan(data_residues_total)] = np.nan

data_residues_hphob[np.isnan(data_residues_total)] = np.nan

#calculate ratios

tmp_div = data_residues_basic + data_residues_hphob

tmp_div[np.isnan(tmp_div)] = 1

data_ratios = data_residues_hphob / tmp_div * 100

global data_ratios_avg

global data_ratios_std

#remove nan values of the weights

weights_nan = np.zeros((nb_rows, 1))

weights_nan_sq = np.zeros((nb_rows, 1))

nb_files = np.ones((nb_rows, 1)) * len(args.xvgfilenames)

tmp_weights_nan = np.zeros((nb_rows, len(args.xvgfilenames)))

for r in range(0, nb_rows):

tmp_weights_nan[r,:] = weights

for f_index in range(0, len(args.xvgfilenames)):

if np.isnan(data_ratios[r,f_index]):

tmp_weights_nan[r,f_index] = 0

nb_files[r,0] -= 1

weights_nan[:,0] = np.nansum(tmp_weights_nan, axis = 1)

weights_nan_sq[:,0] = np.nansum(tmp_weights_nan**2, axis = 1)

weights_nan[weights_nan == 0] = 1

#avg

data_ratios_avg = np.zeros((nb_rows,1))

data_ratios_avg[:,0] = scipy.stats.nanmean(data_ratios * weights * nb_files / weights_nan, axis = 1)

#std

tmp_std = np.zeros((nb_rows, 1))

tmp_std[:,0] = np.nansum(weights * (data_ratios - data_ratios_avg[:,0:1])**2, axis = 1)

tmp_div = np.copy((weights_nan)**2 - weights_nan_sq)

tmp_div[tmp_div == 0] = 1

data_ratios_std = np.sqrt(weights_nan / tmp_div * tmp_std)

#open files

filename_xvg = os.getcwd() + '/' + str(args.output_file) + '.xvg'

output_xvg = open(filename_xvg, 'w')

#general header

output_xvg.write("# [average xvg - written by xvg_average_residues_ratios v" + str(version_nb) + "]\n")

tmp_files = ""

for f in args.xvgfilenames:

tmp_files += "," + str(f)

output_xvg.write("# - files: " + str(tmp_files[1:]) + "\n")

if np.sum(weights) > len(args.xvgfilenames):

output_xvg.write("# -> weight = " + str(np.sum(weights)) + "\n")

#xvg metadata

output_xvg.write("@ title \"Average xvg\"\n")

output_xvg.write("@ xaxis label \"distance from cluster z axis (Angstrom)\"\n")

output_xvg.write("@ yaxis label \"basic / (basic + hydrophobic) in protein residues\"\n")

output_xvg.write("@ autoscale ONREAD xaxes\n")

output_xvg.write("@ TYPE XY\n")

output_xvg.write("@ view 0.15, 0.15, 0.95, 0.85\n")

output_xvg.write("@ legend on\n")

output_xvg.write("@ legend box on\n")

output_xvg.write("@ legend loctype view\n")

output_xvg.write("@ legend 0.98, 0.8\n")

output_xvg.write("@ legend length 2\n")

output_xvg.write("@ s0 legend \"b/(b+h) (avg)\"\n")

output_xvg.write("@ s1 legend \"b/(b+h) (std)\"\n")

#data

for r in range(0, nb_rows):

results = str(distance[r,0])

results += " " + "{:.6e}".format(data_ratios_avg[r,0]) + " " + "{:.6e}".format(data_ratios_std[r,0])

output_xvg.write(results + "\n")

output_xvg.close()