def make_score_vs_rmsd_plot(self, loop):
"""
Create a score vs RMSD plot for the given loop. In fact two plots are
made: one which includes every model and one which includes only the
top 75% best scoring models. Normally the second plot is of more
interest, because it focuses better on the interesting lower-left
region of the plot. The full plots often have outliers that really
scale the score axis.
"""
# This method would be much more concise if it used matplotlib.
if not loop.has_data:
return
tsv_path = os.path.join(loop.latex_dir, 'score_vs_rmsd.tsv')
gnu_path = os.path.join(loop.latex_dir, 'score_vs_rmsd.gnu')
pdf_path_100 = os.path.join(loop.latex_dir, 'score_vs_rmsd_all.pdf')
pdf_path_75 = os.path.join(loop.latex_dir,
'score_vs_rmsd_third_quartile.pdf')
tsv_row = '{0.id}\t{0.rmsd}\t{0.score}\n'
sorted_models = loop.models_sorted_by_score
scores = loop.scores
min_score, max_score = min(scores), max(scores)
third_quartile = numpy.percentile(scores, 75)
native_score = 0 # This isn't stored in the database yet.
# Write score vs RMSD data to a tab-separated value (TSV) file that can
# easily be parsed by gnuplot.
with open(tsv_path, 'w') as file:
file.write('#Model\tLoop_rmsd\tTotal_score\n')
file.write('input_structure\t0.0\t{0}\n'.format(native_score))
# All models
file.write('\n\n')
for model in sorted_models:
file.write(tsv_row.format(model))
# Top X scoring models
file.write('\n\n')
for model in sorted_models[:top_x]:
file.write(tsv_row.format(model))
# Top scoring model
file.write('\n\n')
file.write(tsv_row.format(sorted_models[0]))
# Write the gnuplot script and generate the EPS plots.
gnuplot_script = '''\
set autoscale
set border 31
set tics out
set terminal pdf
set xtics autofreq
set xtics nomirror
set ytics autofreq
set ytics nomirror
set noy2tics
set nox2tics
set style line 1 lt 1 lc rgb "dark-magenta" lw 2
set style line 2 lt 1 lc rgb "{loop.benchmark.color}" lw 2 ps 0.5 pt 7
set style line 3 lt 1 lc rgb "forest-green" lw 2 ps 2 pt 13
set style line 4 lt 1 lc rgb "dark-gray" lw 2 ps 0.5 pt 7
set style line 5 lt 1 lc rgb "black" lw 2 ps 0.8 pt 13
set style line 6 lt 1 lc rgb "black" lw 2
set style line 7 lt 1 lc rgb "dark-gray" lw 2
set style line 8 lt 1 lc rgb "gray" lw 2
set style line 9 lt 2 lc rgb "dark-gray" lw 5
set boxwidth 0.75
set key below right
set xrange [0:]
set encoding iso_8859_1
set title "{loop.pdb_id}: {loop.percent_subangstrom:0.2f}% sub-\305 models"
set xlabel "r.m.s. deviation to crystal loop [\305]"
set arrow from 1, graph 0 to 1, graph 1 ls 9 nohead
set ylabel "Rosetta all-atom score"
set output "{pdf_path_100}"
plot "{tsv_path}" index 1 using ($2):($3) with points ls 2 title "all models" axes x1y1, \\
"{tsv_path}" index 2 using ($2):($3) with points ls 4 title "5 lowest energy models" axes x1y1, \\
"{tsv_path}" index 3 using ($2):($3) with points ls 5 title "top 5 best model" axes x1y1
set yrange [:{third_quartile}]
set output "{pdf_path_75}"
set xrange [0:]
plot "{tsv_path}" index 1 using ($2):($3) with points ls 2 title "75% lowest-scoring models" axes x1y1, \\
"{tsv_path}" index 2 using ($2):($3) with points ls 4 title "5 lowest energy models" axes x1y1, \\
"{tsv_path}" index 3 using ($2):($3) with points ls 5 title "top 5 best model" axes x1y1
'''
with open(gnu_path, 'w') as file:
file.write(gnuplot_script.format(**locals()))
utilities.run_gnuplot(gnu_path, verbose=self.verbose)
return pdf_path_100, pdf_path_75
def make_comparison_plot(self,
distributions,
path_template,
custom_gnuplot_commands,
custom_plot_arguments=''):
"""
Create a plot comparing the same distribution from several different
benchmarks. Examples include the percent subangstrom distribution and
the RMSDs of the lowest scoring predictions. The resulting plot will
have a nicely colored box plot for each benchmark.
Inputs
------
distributions:
A dictionary mapping benchmark objects to some sort of
distribution. The box plot will be created using the distribution
and labeled using the benchmark object.
path_template:
The base file name used to create the TSV, GNU, and EPS files
generated by this method.
custom_gnuplot_commands:
A string containing custom commands to pass to gnuplot immediately
before the 'plot' command. This is meant to be used for doing
things like labeling the axes or adding useful vertical lines.
Outputs
-------
This method creates three files: a TSV file containing the raw data
being plotted, a GNU file containing the gnuplot commands used to
generate the plot, and an EPS file containing the plot itself. The
path to the generated EPS file is also returned.
"""
tsv_path = os.path.join(self.latex_dir, path_template + '.tsv')
gnu_path = os.path.join(self.latex_dir, path_template + '.gnu')
pdf_path = os.path.join(self.latex_dir, path_template + '.pdf')
# Write data to TSV file that can be easily parsed by gnuplot.
boxplot_header = '#' + '\t'.join([
'Protocol', 'x', 'lower', 'first_quartile', 'median',
'third_quartile', 'upper'
]) + '\n'
boxplot_row = '\t'.join([
'{benchmark.name}', '{gnuplot_index}', '{stats.lower_whisker}',
'{stats.first_quartile}', '{stats.median}',
'{stats.third_quartile}', '{stats.upper_whisker}'
]) + '\n'
outlier_header = '#' + '\t'.join(['Protocol', 'x', 'outlier']) + '\n'
outlier_row = '\t'.join(
['{benchmark.name}', '{gnuplot_index}', '{outlier}']) + '\n'
for index, benchmark in enumerate(reversed(distributions)):
distribution = {1: distributions[benchmark]}
boxplots = statistics.tukeyBoxAndWhisker(distribution)
stats, outliers = boxplots[1]
gnuplot_index = index + 1
if not outliers:
outliers = '?'
with open(tsv_path, 'a') as file:
file.write(boxplot_header)
file.write(boxplot_row.format(**locals()))
file.write('\n\n')
file.write(outlier_header)
for outlier in outliers:
file.write(outlier_row.format(**locals()))
file.write('\n\n')
# Generate plot using gnuplot.
x_range = len(distributions) + 1
x_ticks = ', '.join([
'"{0.title}" {1}'.format(benchmark, i + 1)
for i, benchmark in enumerate(reversed(distributions))
])
fig_height = min(1 + len(self), 5)
gnuplot_script = '''\
set autoscale
set border 31
set tics out
set terminal pdf size {fig_height},6
set xtics ({x_ticks}) rotate by -90
set xtics nomirror
set ytics autofreq rotate by -90 center
set ytics nomirror
set noy2tics
set nox2tics
set style line 1 lt 1 lc rgb "dark-magenta" lw 2
set style line 2 lt 1 lc rgb "blue" lw 5 ps 1 pt 7
set style line 3 lt 1 lc rgb "forest-green" lw 2 ps 2 pt 13
set style line 4 lt 1 lc rgb "gold" lw 2 ps 1 pt 7
set style line 5 lt 1 lc rgb "red" lw 2 ps 2 pt 13
set style line 6 lt 1 lc rgb "black" lw 2
set style line 7 lt 1 lc rgb "dark-gray" lw 2
set style line 8 lt 1 lc rgb "gray" lw 2
set style line 9 lt 2 lc rgb "dark-gray" lw 5
set style fill solid 0.5
set boxwidth 0.75
set key below right
set xrange [0:{x_range}]
set encoding iso_8859_1
set notitle
unset xlabel
set yrange [0:]
set output "{pdf_path}"
{custom_gnuplot_commands}
plot {plot_arguments}
'''
plot_template = ', \\\n '.join([
'"{tsv_path}" index {box_plot_index} using 2:4:3:7:6 with candlesticks whiskerbars lt 1 lc rgb "{color}" lw 5 notitle',
'"{tsv_path}" index {box_plot_index} using 2:5:5:5:5 with candlesticks lt 1 lc rgb "black" lw 5 notitle',
'"{tsv_path}" index {outliers_index} using 2:3 with points lt 1 lc rgb "{color}" lw 5 ps 0.5 pt 7 notitle',
])
if not distributions:
raise Exception(
'An error occurred retrieving data from the database.')
plot_arguments = ', \\\n '.join([
plot_template.format(tsv_path=tsv_path,
box_plot_index=2 * i,
outliers_index=2 * i + 1,
color=benchmark.color)
for i, benchmark in enumerate(reversed(distributions))
])
if custom_plot_arguments:
plot_arguments += ', ' + custom_plot_arguments
with open(gnu_path, 'w') as file:
file.write(gnuplot_script.format(**locals()))
utilities.run_gnuplot(gnu_path, verbose=self.verbose)
return pdf_path
def make_summary_box_plots(self, benchmark):
tsv_path = os.path.join(benchmark.latex_dir, 'best_model_dists.tsv')
gnu_path = os.path.join(benchmark.latex_dir, 'best_model_dists.gnu')
pdf_path_rmsd = os.path.join(benchmark.latex_dir,
'best_model_dists_rmsd.pdf')
pdf_path_score = os.path.join(benchmark.latex_dir,
'best_model_dists_score.pdf')
pdf_path_subA = os.path.join(
benchmark.latex_dir, 'best_model_dists_percent_subangstrom.pdf')
# Calculate box plot parameters.
best_top_x_models = benchmark.best_top_x_models
distributions = {
1: [x.rmsd for x in best_top_x_models],
2: [x.score for x in best_top_x_models],
3: benchmark.percents_subangstrom,
}
box_plots = statistics.tukeyBoxAndWhisker(distributions)
# Write box plot data to a tab-separated value (TSV) file that can
# easily be parsed by gnuplot.
with open(tsv_path, 'w') as file:
for x in box_plots:
box_params, outliers = box_plots[x]
file.write('#x\t' + 'lower\t' + 'first_quartile\t' +
'median\t' + 'third_quartile\t' + 'upper\n')
for item in box_params:
file.write('{0}\t'.format(item))
file.write('\n\n\n')
file.write('#x\toutlier\n')
for outlier in outliers:
file.write('{0}\t{1}\n'.format(x, outlier))
if not outliers: file.write('{0}\t?\n'.format(x))
file.write('\n\n')
# Write the gnuplot script and generate the EPS plots.
gnuplot_script = '''\
set autoscale
set border 31
set tics out
set terminal pdf
set size ratio 1
set noxtics
set xrange [0.5:1.5]
set nox2tics
set ytics 1
set ytics nomirror
set noy2tics
set style line 1 lt 1 lc rgb "dark-magenta" lw 2
set style line 2 lt 1 lc rgb "{benchmark.color}" lw 5 pt 7
set style line 3 lt 1 lc rgb "{benchmark.color}" lw 5
set style line 4 lt 1 lc rgb "gold" lw 2
set style line 5 lt 1 lc rgb "red" lw 5 pt 7
set style line 6 lt 1 lc rgb "black" lw 5
set style line 7 lt 1 lc rgb "dark-gray" lw 2
set style line 8 lt 1 lc rgb "gray" lw 2
set style line 9 lt 0 lc rgb "black" lw 5
set boxwidth 0.25
set key tmargin
set title "Best models performance distribution"
set noxlabel
set style fill solid 0.5
set encoding iso_8859_1
set ylabel "r.m.s. deviation to crystal loop [\305]"
set output "{pdf_path_rmsd}"
f(x)=1
plot "{tsv_path}" index 0 using 1:3:2:6:5 with candlesticks whiskerbars ls 2 notitle axes x1y1,\\
"{tsv_path}" index 0 using 1:4:4:4:4 with candlesticks ls 6 notitle,\\
"{tsv_path}" index 1 using 1:2 with points ls 2 ps 0.5 pt 7 notitle,\\
f(x) with lines ls 9 notitle
set ylabel "Rosetta all-atom score"
set xrange [1.5:2.5]
set ytics autofreq
set output "{pdf_path_score}"
plot "{tsv_path}" index 2 using 1:3:2:6:5 with candlesticks whiskerbars ls 5 notitle axes x1y1,\\
"{tsv_path}" index 2 using 1:4:4:4:4 with candlesticks ls 6 notitle,\\
"{tsv_path}" index 3 using 1:2 with points ls 5 ps 0.5 pt 7 notitle
set title "Protocol performance distribution"
set ylabel "Fraction sub-\305 models [%]"
set xrange [2.5:3.5]
set ytics 10
set output "{pdf_path_subA}"
plot "{tsv_path}" index 4 using 1:3:2:6:5 with candlesticks whiskerbars ls 3 notitle axes x1y1,\\
"{tsv_path}" index 4 using 1:4:4:4:4 with candlesticks ls 6 notitle,\\
"{tsv_path}" index 5 using 1:2 with points ls 3 ps 0.5 pt 7 notitle
'''
with open(gnu_path, 'w') as file:
file.write(gnuplot_script.format(**locals()))
# If there are no outliers, gnuplot will produce a warning. This is a
# pretty common occurrence, and I think it's really bad to produce
# warning message for common occurrences. So instead I opt to ignore
# stderr. This is a little dangerous. It would probably be better to
# suppress only the exact warning I know about. But if we were really
# interested in doing things the right way, we would use matplotlib
# instead of gnuplot.
with open(os.devnull) as devnull:
utilities.run_gnuplot(gnu_path,
stderr=devnull,
verbose=self.verbose)
return pdf_path_rmsd, pdf_path_score, pdf_path_subA
def make_rmsd_histogram(self, loop):
"""
Create a smoothed RMSD histogram for the given loop. 100 bins are used
when making the plot, and the smoothing is done by gnuplot.
"""
# This method would be much more concise if it used matplotlib.
if not loop.has_data:
return
tsv_path = os.path.join(loop.latex_dir, 'rmsd_histogram.tsv')
gnu_path = os.path.join(loop.latex_dir, 'rmsd_histogram.gnu')
pdf_path = os.path.join(loop.latex_dir, 'rmsd_histogram.pdf')
# Write histogram data to a tab-separated value (TSV) file that can
# easily be parsed by gnuplot.
num_bins = 100
histogram = statistics.histogram(loop.rmsds, num_bins)
with open(tsv_path, 'w') as file:
file.write('#All models\n')
file.write('#RMSD\tFrequency\n')
for rmsd, count in histogram:
count = num_bins * count / len(loop)
file.write('{0}\t{1}\n'.format(rmsd, count))
# Write the gnuplot script and generate the EPS plot.
gnuplot_script = '''\
set autoscale
set border 31
set tics out
set terminal pdf enhanced color
set xtics autofreq
set xtics nomirror
set ytics autofreq
set ytics nomirror
set noy2tics
set nox2tics
set style line 1 lt 1 lc rgb "dark-magenta" lw 2
set style line 2 lt 1 lc rgb "{loop.benchmark.color}" lw 8 ps 1 pt 7
set style line 3 lt 1 lc rgb "forest-green" lw 2 ps 2 pt 13
set style line 4 lt 1 lc rgb "gold" lw 2 ps 1 pt 7
set style line 5 lt 1 lc rgb "red" lw 2 ps 2 pt 13
set style line 6 lt 1 lc rgb "black" lw 2
set style line 7 lt 1 lc rgb "dark-gray" lw 2
set style line 8 lt 1 lc rgb "gray" lw 2
set style line 9 lt 2 lc rgb "dark-gray" lw 5
set boxwidth 0.75
set key below right
set xrange [0:]
set encoding iso_8859_1
set title "{loop.pdb_id}: {loop.percent_subangstrom:0.2f}% sub-\305 models"
set xlabel "r.m.s. deviation to crystal loop [\305]"
set yrange [0:]
set arrow from 1, graph 0 to 1, graph 1 ls 9 nohead
set ylabel "Fraction of models [%]"
set output "{pdf_path}"
plot "{tsv_path}" index 0 using ($1):($2) smooth bezier with lines ls 2 title "all models" axes x1y1
'''
with open(gnu_path, 'w') as file:
file.write(gnuplot_script.format(**locals()))
utilities.run_gnuplot(gnu_path, verbose=self.verbose)
return pdf_path
def make_score_vs_rmsd_plot(self, loop):
"""
Create a score vs RMSD plot for the given loop. In fact two plots are
made: one which includes every model and one which includes only the
top 75% best scoring models. Normally the second plot is of more
interest, because it focuses better on the interesting lower-left
region of the plot. The full plots often have outliers that really
scale the score axis.
"""
# This method would be much more concise if it used matplotlib.
if not loop.has_data:
return
tsv_path = os.path.join(loop.latex_dir, 'score_vs_rmsd.tsv')
gnu_path = os.path.join(loop.latex_dir, 'score_vs_rmsd.gnu')
pdf_path_100 = os.path.join(loop.latex_dir, 'score_vs_rmsd_all.pdf')
pdf_path_75 = os.path.join(loop.latex_dir, 'score_vs_rmsd_third_quartile.pdf')
tsv_row = '{0.id}\t{0.rmsd}\t{0.score}\n'
sorted_models = loop.models_sorted_by_score
scores = loop.scores
min_score, max_score = min(scores), max(scores)
third_quartile = numpy.percentile(scores, 75)
native_score = 0 # This isn't stored in the database yet.
# Write score vs RMSD data to a tab-separated value (TSV) file that can
# easily be parsed by gnuplot.
with open(tsv_path, 'w') as file:
file.write('#Model\tLoop_rmsd\tTotal_score\n')
file.write('input_structure\t0.0\t{0}\n'.format(native_score))
# All models
file.write('\n\n')
for model in sorted_models:
file.write(tsv_row.format(model))
# Top X scoring models
file.write('\n\n')
for model in sorted_models[:top_x]:
file.write(tsv_row.format(model))
# Top scoring model
file.write('\n\n')
file.write(tsv_row.format(sorted_models[0]))
# Write the gnuplot script and generate the EPS plots.
gnuplot_script = '''\
set autoscale
set border 31
set tics out
set terminal pdf
set xtics autofreq
set xtics nomirror
set ytics autofreq
set ytics nomirror
set noy2tics
set nox2tics
set style line 1 lt 1 lc rgb "dark-magenta" lw 2
set style line 2 lt 1 lc rgb "{loop.benchmark.color}" lw 2 ps 0.5 pt 7
set style line 3 lt 1 lc rgb "forest-green" lw 2 ps 2 pt 13
set style line 4 lt 1 lc rgb "dark-gray" lw 2 ps 0.5 pt 7
set style line 5 lt 1 lc rgb "black" lw 2 ps 0.8 pt 13
set style line 6 lt 1 lc rgb "black" lw 2
set style line 7 lt 1 lc rgb "dark-gray" lw 2
set style line 8 lt 1 lc rgb "gray" lw 2
set style line 9 lt 2 lc rgb "dark-gray" lw 5
set boxwidth 0.75
set key below right
set xrange [0:]
set encoding iso_8859_1
set title "{loop.pdb_id}: {loop.percent_subangstrom:0.2f}% sub-\305 models"
set xlabel "r.m.s. deviation to crystal loop [\305]"
set arrow from 1, graph 0 to 1, graph 1 ls 9 nohead
set ylabel "Rosetta all-atom score"
set output "{pdf_path_100}"
plot "{tsv_path}" index 1 using ($2):($3) with points ls 2 title "all models" axes x1y1, \\
"{tsv_path}" index 2 using ($2):($3) with points ls 4 title "5 lowest energy models" axes x1y1, \\
"{tsv_path}" index 3 using ($2):($3) with points ls 5 title "top 5 best model" axes x1y1
set yrange [:{third_quartile}]
set output "{pdf_path_75}"
set xrange [0:]
plot "{tsv_path}" index 1 using ($2):($3) with points ls 2 title "75% lowest-scoring models" axes x1y1, \\
"{tsv_path}" index 2 using ($2):($3) with points ls 4 title "5 lowest energy models" axes x1y1, \\
"{tsv_path}" index 3 using ($2):($3) with points ls 5 title "top 5 best model" axes x1y1
'''
with open(gnu_path, 'w') as file:
file.write(gnuplot_script.format(**locals()))
utilities.run_gnuplot(gnu_path, verbose=self.verbose)
return pdf_path_100, pdf_path_75
def make_summary_box_plots(self, benchmark):
tsv_path = os.path.join(benchmark.latex_dir, 'best_model_dists.tsv')
gnu_path = os.path.join(benchmark.latex_dir, 'best_model_dists.gnu')
pdf_path_rmsd = os.path.join(benchmark.latex_dir, 'best_model_dists_rmsd.pdf')
pdf_path_score = os.path.join(benchmark.latex_dir, 'best_model_dists_score.pdf')
pdf_path_subA = os.path.join(benchmark.latex_dir, 'best_model_dists_percent_subangstrom.pdf')
# Calculate box plot parameters.
best_top_x_models = benchmark.best_top_x_models
distributions = {
1: [x.rmsd for x in best_top_x_models],
2: [x.score for x in best_top_x_models],
3: benchmark.percents_subangstrom,
}
box_plots = statistics.tukeyBoxAndWhisker(distributions)
# Write box plot data to a tab-separated value (TSV) file that can
# easily be parsed by gnuplot.
with open(tsv_path, 'w') as file:
for x in box_plots:
box_params, outliers = box_plots[x]
file.write('#x\t'+'lower\t'+'first_quartile\t'+'median\t'+'third_quartile\t'+'upper\n')
for item in box_params: file.write('{0}\t'.format(item))
file.write('\n\n\n')
file.write('#x\toutlier\n')
for outlier in outliers: file.write('{0}\t{1}\n'.format(x, outlier))
if not outliers: file.write('{0}\t?\n'.format(x))
file.write('\n\n')
# Write the gnuplot script and generate the EPS plots.
gnuplot_script = '''\
set autoscale
set border 31
set tics out
set terminal pdf
set size ratio 1
set noxtics
set xrange [0.5:1.5]
set nox2tics
set ytics 1
set ytics nomirror
set noy2tics
set style line 1 lt 1 lc rgb "dark-magenta" lw 2
set style line 2 lt 1 lc rgb "{benchmark.color}" lw 5 pt 7
set style line 3 lt 1 lc rgb "{benchmark.color}" lw 5
set style line 4 lt 1 lc rgb "gold" lw 2
set style line 5 lt 1 lc rgb "red" lw 5 pt 7
set style line 6 lt 1 lc rgb "black" lw 5
set style line 7 lt 1 lc rgb "dark-gray" lw 2
set style line 8 lt 1 lc rgb "gray" lw 2
set style line 9 lt 0 lc rgb "black" lw 5
set boxwidth 0.25
set key tmargin
set title "Best models performance distribution"
set noxlabel
set style fill solid 0.5
set encoding iso_8859_1
set ylabel "r.m.s. deviation to crystal loop [\305]"
set output "{pdf_path_rmsd}"
f(x)=1
plot "{tsv_path}" index 0 using 1:3:2:6:5 with candlesticks whiskerbars ls 2 notitle axes x1y1,\\
"{tsv_path}" index 0 using 1:4:4:4:4 with candlesticks ls 6 notitle,\\
"{tsv_path}" index 1 using 1:2 with points ls 2 ps 0.5 pt 7 notitle,\\
f(x) with lines ls 9 notitle
set ylabel "Rosetta all-atom score"
set xrange [1.5:2.5]
set ytics autofreq
set output "{pdf_path_score}"
plot "{tsv_path}" index 2 using 1:3:2:6:5 with candlesticks whiskerbars ls 5 notitle axes x1y1,\\
"{tsv_path}" index 2 using 1:4:4:4:4 with candlesticks ls 6 notitle,\\
"{tsv_path}" index 3 using 1:2 with points ls 5 ps 0.5 pt 7 notitle
set title "Protocol performance distribution"
set ylabel "Fraction sub-\305 models [%]"
set xrange [2.5:3.5]
set ytics 10
set output "{pdf_path_subA}"
plot "{tsv_path}" index 4 using 1:3:2:6:5 with candlesticks whiskerbars ls 3 notitle axes x1y1,\\
"{tsv_path}" index 4 using 1:4:4:4:4 with candlesticks ls 6 notitle,\\
"{tsv_path}" index 5 using 1:2 with points ls 3 ps 0.5 pt 7 notitle
'''
with open(gnu_path, 'w') as file:
file.write(gnuplot_script.format(**locals()))
# If there are no outliers, gnuplot will produce a warning. This is a
# pretty common occurrence, and I think it's really bad to produce
# warning message for common occurrences. So instead I opt to ignore
# stderr. This is a little dangerous. It would probably be better to
# suppress only the exact warning I know about. But if we were really
# interested in doing things the right way, we would use matplotlib
# instead of gnuplot.
with open(os.devnull) as devnull:
utilities.run_gnuplot(gnu_path, stderr=devnull, verbose=self.verbose)
return pdf_path_rmsd, pdf_path_score, pdf_path_subA
def make_comparison_plot(self, distributions, path_template,
custom_gnuplot_commands, custom_plot_arguments=''):
"""
Create a plot comparing the same distribution from several different
benchmarks. Examples include the percent subangstrom distribution and
the RMSDs of the lowest scoring predictions. The resulting plot will
have a nicely colored box plot for each benchmark.
Inputs
------
distributions:
A dictionary mapping benchmark objects to some sort of
distribution. The box plot will be created using the distribution
and labeled using the benchmark object.
path_template:
The base file name used to create the TSV, GNU, and EPS files
generated by this method.
custom_gnuplot_commands:
A string containing custom commands to pass to gnuplot immediately
before the 'plot' command. This is meant to be used for doing
things like labeling the axes or adding useful vertical lines.
Outputs
-------
This method creates three files: a TSV file containing the raw data
being plotted, a GNU file containing the gnuplot commands used to
generate the plot, and an EPS file containing the plot itself. The
path to the generated EPS file is also returned.
"""
tsv_path = os.path.join(self.latex_dir, path_template + '.tsv')
gnu_path = os.path.join(self.latex_dir, path_template + '.gnu')
pdf_path = os.path.join(self.latex_dir, path_template + '.pdf')
# Write data to TSV file that can be easily parsed by gnuplot.
boxplot_header = '#' + '\t'.join([
'Protocol',
'x',
'lower',
'first_quartile',
'median',
'third_quartile',
'upper']) + '\n'
boxplot_row = '\t'.join([
'{benchmark.name}',
'{gnuplot_index}',
'{stats.lower_whisker}',
'{stats.first_quartile}',
'{stats.median}',
'{stats.third_quartile}',
'{stats.upper_whisker}']) + '\n'
outlier_header = '#' + '\t'.join([
'Protocol',
'x',
'outlier']) + '\n'
outlier_row = '\t'.join([
'{benchmark.name}',
'{gnuplot_index}',
'{outlier}']) + '\n'
for index, benchmark in enumerate(reversed(distributions)):
distribution = {1: distributions[benchmark]}
boxplots = statistics.tukeyBoxAndWhisker(distribution)
stats, outliers = boxplots[1]
gnuplot_index = index + 1
if not outliers:
outliers = '?'
with open(tsv_path, 'a') as file:
file.write(boxplot_header)
file.write(boxplot_row.format(**locals()))
file.write('\n\n')
file.write(outlier_header)
for outlier in outliers:
file.write(outlier_row.format(**locals()))
file.write('\n\n')
# Generate plot using gnuplot.
x_range = len(distributions) + 1
x_ticks = ', '.join([
'"{0.title}" {1}'.format(benchmark, i+1)
for i, benchmark in enumerate(reversed(distributions))
])
fig_height = min(1 + len(self), 5)
gnuplot_script = '''\
set autoscale
set border 31
set tics out
set terminal pdf size {fig_height},6
set xtics ({x_ticks}) rotate by -90
set xtics nomirror
set ytics autofreq rotate by -90 center
set ytics nomirror
set noy2tics
set nox2tics
set style line 1 lt 1 lc rgb "dark-magenta" lw 2
set style line 2 lt 1 lc rgb "blue" lw 5 ps 1 pt 7
set style line 3 lt 1 lc rgb "forest-green" lw 2 ps 2 pt 13
set style line 4 lt 1 lc rgb "gold" lw 2 ps 1 pt 7
set style line 5 lt 1 lc rgb "red" lw 2 ps 2 pt 13
set style line 6 lt 1 lc rgb "black" lw 2
set style line 7 lt 1 lc rgb "dark-gray" lw 2
set style line 8 lt 1 lc rgb "gray" lw 2
set style line 9 lt 2 lc rgb "dark-gray" lw 5
set style fill solid 0.5
set boxwidth 0.75
set key below right
set xrange [0:{x_range}]
set encoding iso_8859_1
set notitle
unset xlabel
set yrange [0:]
set output "{pdf_path}"
{custom_gnuplot_commands}
plot {plot_arguments}
'''
plot_template = ', \\\n '.join([
'"{tsv_path}" index {box_plot_index} using 2:4:3:7:6 with candlesticks whiskerbars lt 1 lc rgb "{color}" lw 5 notitle',
'"{tsv_path}" index {box_plot_index} using 2:5:5:5:5 with candlesticks lt 1 lc rgb "black" lw 5 notitle',
'"{tsv_path}" index {outliers_index} using 2:3 with points lt 1 lc rgb "{color}" lw 5 ps 0.5 pt 7 notitle',
])
if not distributions:
raise Exception('An error occurred retrieving data from the database.')
plot_arguments = ', \\\n '.join([
plot_template.format(
tsv_path=tsv_path,
box_plot_index=2*i,
outliers_index=2*i+1,
color=benchmark.color)
for i, benchmark in enumerate(reversed(distributions))
])
if custom_plot_arguments:
plot_arguments += ', ' + custom_plot_arguments
with open(gnu_path, 'w') as file:
file.write(gnuplot_script.format(**locals()))
utilities.run_gnuplot(gnu_path, verbose=self.verbose)
return pdf_path
def make_rmsd_histogram(self, loop):
"""
Create a smoothed RMSD histogram for the given loop. 100 bins are used
when making the plot, and the smoothing is done by gnuplot.
"""
# This method would be much more concise if it used matplotlib.
if not loop.has_data:
return
tsv_path = os.path.join(loop.latex_dir, 'rmsd_histogram.tsv')
gnu_path = os.path.join(loop.latex_dir, 'rmsd_histogram.gnu')
pdf_path = os.path.join(loop.latex_dir, 'rmsd_histogram.pdf')
# Write histogram data to a tab-separated value (TSV) file that can
# easily be parsed by gnuplot.
num_bins = 100
histogram = statistics.histogram(loop.rmsds, num_bins)
with open(tsv_path, 'w') as file:
file.write('#All models\n')
file.write('#RMSD\tFrequency\n')
for rmsd, count in histogram:
count = num_bins * count / len(loop)
file.write('{0}\t{1}\n'.format(rmsd, count))
# Write the gnuplot script and generate the EPS plot.
gnuplot_script='''\
set autoscale
set border 31
set tics out
set terminal pdf enhanced color
set xtics autofreq
set xtics nomirror
set ytics autofreq
set ytics nomirror
set noy2tics
set nox2tics
set style line 1 lt 1 lc rgb "dark-magenta" lw 2
set style line 2 lt 1 lc rgb "{loop.benchmark.color}" lw 8 ps 1 pt 7
set style line 3 lt 1 lc rgb "forest-green" lw 2 ps 2 pt 13
set style line 4 lt 1 lc rgb "gold" lw 2 ps 1 pt 7
set style line 5 lt 1 lc rgb "red" lw 2 ps 2 pt 13
set style line 6 lt 1 lc rgb "black" lw 2
set style line 7 lt 1 lc rgb "dark-gray" lw 2
set style line 8 lt 1 lc rgb "gray" lw 2
set style line 9 lt 2 lc rgb "dark-gray" lw 5
set boxwidth 0.75
set key below right
set xrange [0:]
set encoding iso_8859_1
set title "{loop.pdb_id}: {loop.percent_subangstrom:0.2f}% sub-\305 models"
set xlabel "r.m.s. deviation to crystal loop [\305]"
set yrange [0:]
set arrow from 1, graph 0 to 1, graph 1 ls 9 nohead
set ylabel "Fraction of models [%]"
set output "{pdf_path}"
plot "{tsv_path}" index 0 using ($1):($2) smooth bezier with lines ls 2 title "all models" axes x1y1
'''
with open(gnu_path, 'w') as file:
file.write(gnuplot_script.format(**locals()))
utilities.run_gnuplot(gnu_path, verbose=self.verbose)
return pdf_path