Index: doc/theses/mike_brooks_MMath/plots/common.py =================================================================== --- doc/theses/mike_brooks_MMath/plots/common.py (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/common.py (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,113 @@ +import pandas as pd +import numpy as np +import sys +import os + +def parseTestCorpus(dt): + dt[['test-slug', + 'sut-platform', + 'operation', + 'sut-cfa-level', + 'sut-cfa-sharing', + 'op-alloc']] = dt['test'].str.strip().str.split('-', expand=True) + dt['sut'] = dt[['sut-platform', + 'sut-cfa-level', + 'sut-cfa-sharing', + 'op-alloc']].agg('-'.join, axis=1) + + dt[['corpus-basename', + 'corpus-ext']] = dt['corpus'].str.strip().str.split('.', expand=True) + dt[['corpus-slug', + 'corpus-nstrs', + 'corpus-meanlen', + 'corpus-runid']] = dt['corpus-basename'].str.strip().str.split('-', expand=True) + dt["corpus-nstrs"] = pd.to_numeric(dt["corpus-nstrs"]) + dt["corpus-meanlen"] = pd.to_numeric(dt["corpus-meanlen"]) + dt["corpus-runid"] = pd.to_numeric(dt["corpus-runid"]) + +def loadParseTimingData( infileLocal, xClasNames=[], xClasDtypes={}, xFactNames=[], xFactDtypes={} ): + + infile = os.path.dirname(os.path.abspath(__file__)) + '/../benchmarks/string/' + infileLocal + + timings = pd.read_csv( + infile, + names=['test', 'corpus'] + xClasNames + [ 'concatsPerReset', 'corpusItemCount', 'corpusMeanLenChars', 'concatDoneActualCount', 'execTimeActualSec'] + xFactNames, + dtype={**xClasDtypes, **xFactDtypes, **{ + 'test': str, + 'corpus': str, + 'concatsPerReset': 'Int64', # allows missing; https://stackoverflow.com/a/70626154 + 'corpusItemCount': np.int64, + 'corpusMeanLenChars': np.float64, + 'concatDoneActualCount': np.int64, + 'execTimeActualSec': np.float64, + 'Q': np.int64}}, + + na_values=['xxx'], + ) + # print(timings.head()) + + + # project: parse executable and corpus names + + parseTestCorpus(timings) + + # project: calculate fact + + timings['op-duration-s'] = timings['execTimeActualSec'] / timings['concatDoneActualCount'] + timings['op-duration-ns'] = timings['op-duration-s'] * 1000 * 1000 * 1000 + + return timings + +def loadParseSizingData( infileLocal, xClasNames=[], xClasDtypes={}, xFactNames=[], xFactDtypes={} ): + + infile = os.path.dirname(os.path.abspath(__file__)) + '/../benchmarks/string/' + infileLocal + + sizings = pd.read_csv( + infile, + sep=' ', + names=['test', 'corpus'] + xClasNames + ['ppid', 'pid', 'malloc_count', 'free_count', + 'calloc_count', 'realloc_count', + 'requsted_mem(B)', 'current_req_mem(B)', 'hw_cur_req_mem(B)', 'text', 'heap', 'mmap_so', 'mmap', + 'stack', 'vvar', 'vdso', 'vsyscall', 'unfigured', 'total_dynamic', + 'epoch_timestamp(ms)'] + xFactNames, + dtype={**xClasDtypes, **xFactDtypes, **{ + 'test': str, + 'corpus': str, + 'ppid': np.int64, 'pid': np.int64, 'malloc_count': np.int64, 'free_count': np.int64, + 'calloc_count': np.int64, 'realloc_count': np.int64, + 'requsted_mem(B)': np.int64, 'current_req_mem(B)': np.int64, + 'hw_cur_req_mem(B)': np.int64, 'text': np.int64, 'heap': np.int64, + 'mmap_so': np.int64, 'mmap': np.int64, + 'stack': np.int64, 'vvar': np.int64, 'vdso': np.int64, 'vsyscall': np.int64, 'unfigured': np.int64, 'total_dynamic': np.int64, + 'epoch_timestamp(ms)': np.int64}} + ) + + parseTestCorpus(sizings) + + return sizings + + +def loadParseAttribData( infileLocal ): + + infile = os.path.dirname(os.path.abspath(__file__)) + '/../benchmarks/string/' + infileLocal + + attribs = pd.read_csv( + infile, + sep=' ', + names=[ + "test", "corpus", "expansion", "category", "samples_in_category", "total_samples", + "fraction", "sources"], + dtype={ + "test": str, + "corpus": str, + "expansion": np.float64, + "category": str, + "samples_in_category": np.int64, + "total_samples": np.int64, + "fraction": np.float64, + "sources": str} + ) + + parseTestCorpus(attribs) + + return attribs Index: doc/theses/mike_brooks_MMath/plots/string-allocn-attrib.py =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-allocn-attrib.py (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-allocn-attrib.py (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,56 @@ +# Read thesis-append-pbv.csv +# Output for string-graph-peq-sharing.dat + +# Project details +# Filter operation=peq +# Split "series" goups of sut; only those in the "pretty" list +# Assert one row per string-length +# output: +# string-len op-duration +# in chunks, each headed by pertty(sut) + +import pandas as pd +import numpy as np +import os +import sys + +sys.path.insert(0, os.path.dirname(__file__)) +from common import * + +sizes_i_want = [50, 200] # [20, 50, 100, 200] + +# assume CFA threshold only run at default value + +cfatimings = loadParseTimingData('result-allocate-speed-cfa.csv', + xClasNames=['expansion'], xClasDtypes={'expansion':'Float64'}, + xFactNames=['topIters'], xFactDtypes={'topIters':np.int64}) + +cfaattribs = loadParseAttribData('result-allocate-attrib-cfa.ssv') + +stltimings = loadParseTimingData('result-allocate-speed-stl.csv', + xClasNames=['expansion'], xClasDtypes={'expansion':'Float64'}, + xFactNames=['topIters'], xFactDtypes={'topIters':np.int64}) + +stlattribs = loadParseAttribData('result-allocate-attrib-stl.ssv') + +timings = pd.concat([cfatimings, stltimings]) +attribs = pd.concat([cfaattribs, stlattribs]) + +combined = pd.merge( + left=timings[['sut-platform', 'corpus-meanlen','expansion', 'op-duration-ns']], + right=attribs[['sut-platform', 'corpus-meanlen','expansion', 'category', 'fraction']], + on=['sut-platform', 'corpus-meanlen','expansion'] +) + +combined['cat-duration-ns'] = combined['op-duration-ns'] * combined['fraction'] +combined.drop(columns=['expansion', 'op-duration-ns', 'fraction'], inplace=True) + +pvt = combined.pivot( columns='category', values='cat-duration-ns', index=['corpus-meanlen', 'sut-platform'] ) + +desired_dcol_order = ["ctor-dtor", "gc", "malloc-free", "text-import", "harness-leaf", "other"] +pvt = pvt[desired_dcol_order] + +filtered = pvt.loc[pvt.index.get_level_values('corpus-meanlen').isin(sizes_i_want)] + +print(filtered.to_csv(header=True, index=True, sep='\t', na_rep="0")) + Index: doc/theses/mike_brooks_MMath/plots/string-allocn.d =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-allocn.d (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-allocn.d (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,8 @@ +plots/string-allocn.gp.INPUTS: build/plot-string-allocn.dat +plots/string-allocn.gp.INPUTS: build/plot-string-allocn-attrib.dat +plots/string-allocn.py.INPUTS: benchmarks/string/result-allocate-space-cfa.ssv +plots/string-allocn.py.INPUTS: benchmarks/string/result-allocate-space-stl.ssv +plots/string-allocn.py.INPUTS: benchmarks/string/result-allocate-speed-cfa.csv +plots/string-allocn.py.INPUTS: benchmarks/string/result-allocate-speed-stl.csv +plots/string-allocn-attrib.py.INPUTS: benchmarks/string/result-allocate-attrib-stl.ssv +plots/string-allocn-attrib.py.INPUTS: benchmarks/string/result-allocate-attrib-cfa.ssv Index: doc/theses/mike_brooks_MMath/plots/string-allocn.gp =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-allocn.gp (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-allocn.gp (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,108 @@ +set terminal pdf color enhanced size 6.0in,4.0in font "Times,17" +#set terminal postscript portrait enhanced size 7.5, 10. color solid 9.5; +#set terminal wxt size 950,1250 + +INDIR="build" +OUTDIR="build" + +SCALE=1024 + +# common to all +set macros +set output OUTDIR."/plot-string-allocn.pdf" +set multiplot layout 1, 3 ; +set key outside top center horizontal +set grid + +# common to only first two graphs +set logscale x 2 +#set mxtics 3 # 3 steps within each doubling (e.g. 3 steps of of 32 between 32 and 128 => tick on 32s in there) +set xlabel "Heap Used (B)" +set logscale y 10 + +# +# CFA multisize +# + +set ylabel "Duration (ns)" offset 2,0 +set yrange[35:125] +set ytics add (40, 50, 60, 70, 80, 90, 110, 120) + +set xrange[32:4096] +set xtics rotate by -90 +set xtics ("" 32, "64 k" 64, "" 128, "256 k" 256, "" 512, "1 M" 1024, "" 2048, "4 M" 4096) + +# First each curve, then each default-expansion point +plot INDIR."/plot-string-allocn.dat" \ + i 0 using ($3/SCALE):2 title columnheader(1) with linespoints lt rgb "blue" pt 2 ps 1 lw 1, \ + '' i 1 using ($3/SCALE):2 title columnheader(1) with linespoints lt rgb "red" pt 3 ps 1 lw 1, \ + '' i 2 using ($3/SCALE):2 title columnheader(1) with linespoints lt rgb "brown" pt 8 ps 1 lw 1, \ + '' i 3 using ($3/SCALE):2 title columnheader(1) with linespoints lt rgb "black" pt 10 ps 1 lw 1, \ + '' i 4 using ($3/SCALE):2 title columnheader(1) with linespoints lt rgb "magenta" pt 12 ps 1 lw 1, \ + '' i 0 using ( ($4 == 1) ? ($3/SCALE) : 1/0 ):2 notitle with points lt rgb "blue" pt 66 ps 2, \ + '' i 1 using ( ($4 == 1) ? ($3/SCALE) : 1/0 ):2 notitle with points lt rgb "red" pt 66 ps 2, \ + '' i 2 using ( ($4 == 1) ? ($3/SCALE) : 1/0 ):2 notitle with points lt rgb "brown" pt 66 ps 2, \ + '' i 3 using ( ($4 == 1) ? ($3/SCALE) : 1/0 ):2 notitle with points lt rgb "black" pt 66 ps 2, \ + '' i 4 using ( ($4 == 1) ? ($3/SCALE) : 1/0 ):2 notitle with points lt rgb "magenta" pt 66 ps 2 + +unset ylabel + +unset xtics +unset ytics +unset yrange +unset xrange + + +# +# STL comparison +# + + +set yrange[40:85] +set ytics add (40, 45, 50, 55, 60, 65, 70, 75, 80, 85) +set mytics 90 # 90 steps within each decade (e.g. 90 steps of of 1 between 10 and 100 => tick on 1s in there) + +set xrange[64:4096] +set xtics rotate by -90 +set xtics ("64 k" 64, "128 k" 128, "256 k" 256, "512 k" 512, "1 M" 1024, "2 M" 2048, "4 M" 4096) + +# skullduggeries: +# hardcoding chunk index and assuming data in the chunk (by hardcoding only the stl series title) +# series order is meaningless but important: achieves z-order readability and legend order acceptability + +plot INDIR."/plot-string-allocn.dat" \ + i 8 using ($3/SCALE) :2 title "tradeoff" with lines lt rgb "#77000000" dt (2,2) lw 8, \ + '' i 1 using ($3/SCALE) :2 title columnheader(1) with linespoints lt rgb "red" pt 3 ps 1 lw 1, \ + '' i 6 using ( (strcol(4) eq "cfa") ? ($3/SCALE) : 1/0 ):2 notitle with points lt rgb "red" pt 66 ps 2, \ + '' i 6 using ($3/SCALE) :2 notitle with lines lt rgb "#77000000" dt (2,2) lw 8, \ + '' i 6 using ( (strcol(4) eq "stl") ? ($3/SCALE) : 1/0 ):2 title "stl, len=50" with points lt rgb "red" pt 5 ps 1, \ + '' i 3 using ($3/SCALE) :2 title columnheader(1) with linespoints lt rgb "black" pt 10 ps 1 lw 1, \ + '' i 8 using ( (strcol(4) eq "stl") ? ($3/SCALE) : 1/0 ):2 title "stl, len=200" with points lt rgb "black" pt 5 ps 1, \ + '' i 8 using ( (strcol(4) eq "cfa") ? ($3/SCALE) : 1/0 ):2 notitle with points lt rgb "black" pt 66 ps 2 + + +unset mytics +unset ytics +unset yrange +unset xrange + + + +# common to first two graphs +unset logscale +unset xlabel +unset mxtics + +# +# Attribution +# + +set style data histogram +set ytics auto +set style histogram clustered gap 1 rowstacked +set style fill solid border -1 +set boxwidth 0.8 +set xtics rotate by -45 + +plot for [col=3:8] \ + INDIR.'/plot-string-allocn-attrib.dat' using col:xticlabels(stringcolumn(2).", len=".stringcolumn(1)) index 0 title columnheader(col) Index: doc/theses/mike_brooks_MMath/plots/string-allocn.py =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-allocn.py (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-allocn.py (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,97 @@ +# Read thesis-append-pbv.csv +# Output for string-graph-peq-sharing.dat + +# Project details +# Filter operation=peq +# Split "series" goups of sut; only those in the "pretty" list +# Assert one row per string-length +# output: +# string-len op-duration +# in chunks, each headed by pertty(sut) + +import pandas as pd +import numpy as np +import os +import sys + +sys.path.insert(0, os.path.dirname(__file__)) +from common import * + +# re: apparent cherrypicking +# The system's response to the liveness threshold is not smooth. +# The system only uses the threshold to decide whether it will double the text heap again or not. +# The system's speed for a given string size in a given amount of memory is not affected by the specific value of the liveness threshold. +# Goals with this selection are +# - showing one speed result per +# - cropping diminishing or negative returns for large memory sizes +# - diminishing is obvious, already shown past chosen sweet spot in this selection +# - negative caused by overflowing llc, not relevant to sting impl +favSizes = {20:[-1.0, 0.05, 0.1, 0.2, 0.5, 0.9], + 50:[-1.0, 0.05, 0.1, 0.2, 0.5, 0.9], + 100:[-1.0, 0.1, 0.2, 0.5, 0.9], + 200:[-1.0, 0.1, 0.2, 0.5, 0.9], + 500:[-1.0, 0.4, 0.9, 0.98]} + +defaultExpansions = [-1, 0.2] + +cfatimings = loadParseTimingData('result-allocate-speed-cfa.csv', + xClasNames=['expansion'], xClasDtypes={'expansion':'Float64'}, + xFactNames=['topIters'], xFactDtypes={'topIters':np.int64}) + +cfasizings = loadParseSizingData('result-allocate-space-cfa.ssv', xClasNames=['expansion'], xClasDtypes={'expansion':'Float64'}) + +stltimings = loadParseTimingData('result-allocate-speed-stl.csv', + xClasNames=['expansion'], xClasDtypes={'expansion':'Float64'}, + xFactNames=['topIters'], xFactDtypes={'topIters':np.int64}) + +stlsizings = loadParseSizingData('result-allocate-space-stl.ssv', xClasNames=['expansion'], xClasDtypes={'expansion':'Float64'}) + +timings = pd.concat([cfatimings, stltimings]) +sizings = pd.concat([cfasizings, stlsizings]) + +combined = pd.merge( + left=timings, + right=sizings[['sut', 'corpus','expansion','hw_cur_req_mem(B)']], + on=['sut', 'corpus','expansion'] +) + +combined['is-default'] = np.isin(combined['expansion'], defaultExpansions).astype(int) + +# print ('!!') +# print(combined) + + +# Emit + +# First, for the CFA curves +sut = "cfa" +sutGroup = combined.groupby('sut-platform').get_group(sut) + +groupedSize = sutGroup.groupby('corpus-meanlen') + +for sz, szgroup in groupedSize: + + if sz in favSizes.keys(): + szgroup_sorted = szgroup.sort_values(by='expansion') + + print('"{sut}, len={len}"'.format(sut=sut, len=sz)) + # print(szgroup_sorted) ## + # print(szgroup_sorted['expansion'], 'isin', favSizes[sz]) ## + favoured = szgroup_sorted.loc[szgroup_sorted['expansion'].isin(favSizes[sz])] + # print('!') ## + # print(favoured) ## + text = favoured[['expansion', 'op-duration-ns', 'hw_cur_req_mem(B)', 'is-default']].to_csv(header=False, index=False, sep='\t') + print(text) + print() + +# Again, for the STL-comparisons, default expansion only + +atDefaults = combined.groupby('is-default').get_group(1) + +for sz, szgroup in atDefaults.groupby('corpus-meanlen'): + + if sz in favSizes.keys(): + print(sz) + text = szgroup[['expansion', 'op-duration-ns', 'hw_cur_req_mem(B)', 'sut-platform']].to_csv(header=False, index=False, sep='\t') + print(text) + print() Index: doc/theses/mike_brooks_MMath/plots/string-pbv-fixcorp.py =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-pbv-fixcorp.py (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-pbv-fixcorp.py (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,48 @@ +# Read thesis-append-pbv.csv +# Output for string-graph-peq-sharing.dat + +# Project details +# Filter operation=peq +# Split "series" goups of sut; only those in the "pretty" list +# Assert one row per string-length +# output: +# string-len op-duration +# in chunks, each headed by pertty(sut) + +import pandas as pd +import numpy as np +import os +import sys + +sys.path.insert(0, os.path.dirname(__file__)) +from common import * + +prettyFieldNames = { + "cfa-ll-share-na": "{/Helvetica=15 C{/Symbol \\42}} share", + "stl-na-na-na": "STL", +} + +timings = loadParseTimingData('result-append-pbv.csv') + + +# Filter operation=pbv, corpus=1-*-1 + +timings = timings.groupby('operation').get_group('pbv') +timings = timings.groupby('corpus-nstrs').get_group(1) +timings = timings.groupby('corpus-runid').get_group(1) + + +# Emit in groups + +groupedSut = timings.groupby('sut') + +for sut, sgroup in groupedSut: + + if sut in prettyFieldNames: + + sgroup_sorted = sgroup.sort_values(by='corpus-meanlen') + + print('"{header}"'.format(header=prettyFieldNames[sut])) + text = sgroup_sorted[['corpus-meanlen', 'op-duration-ns']].to_csv(header=False, index=False, sep='\t') + print(text) + print() Index: doc/theses/mike_brooks_MMath/plots/string-pbv-varcorp.py =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-pbv-varcorp.py (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-pbv-varcorp.py (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,48 @@ +# Read thesis-append-pbv.csv +# Output for string-graph-peq-sharing.dat + +# Project details +# Filter operation=peq +# Split "series" goups of sut; only those in the "pretty" list +# Assert one row per string-length +# output: +# string-len op-duration +# in chunks, each headed by pertty(sut) + +import pandas as pd +import numpy as np +import os +import sys + +sys.path.insert(0, os.path.dirname(__file__)) +from common import * + +prettyFieldNames = { + "cfa-ll-share-na": "{/Helvetica=15 C{/Symbol \\42}} share", + "stl-na-na-na": "STL", +} + +timings = loadParseTimingData('result-append-pbv.csv') + + +# Filter operation=pbv, corpus=100-*-1 + +timings = timings.groupby('operation').get_group('pbv') +timings = timings.groupby('corpus-nstrs').get_group(100) +timings = timings.groupby('corpus-runid').get_group(1) + + +# Emit in groups + +groupedSut = timings.groupby('sut') + +for sut, sgroup in groupedSut: + + if sut in prettyFieldNames: + + sgroup_sorted = sgroup.sort_values(by='corpus-meanlen') + + print('"{header}"'.format(header=prettyFieldNames[sut])) + text = sgroup_sorted[['corpus-meanlen', 'op-duration-ns']].to_csv(header=False, index=False, sep='\t') + print(text) + print() Index: doc/theses/mike_brooks_MMath/plots/string-pbv.d =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-pbv.d (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-pbv.d (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,3 @@ +plots/string-pbv.gp.INPUTS: build/plot-string-pbv-varcorp.dat build/plot-string-pbv-fixcorp.dat | build +plots/string-pbv-varcorp.py.INPUTS: benchmarks/string/result-append-pbv.csv +plots/string-pbv-fixcorp.py.INPUTS: benchmarks/string/result-append-pbv.csv Index: doc/theses/mike_brooks_MMath/plots/string-pbv.gp =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-pbv.gp (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-pbv.gp (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,36 @@ +set terminal pdf color enhanced size 6.0in,3.0in font "Times,17" +#set terminal postscript portrait enhanced size 7.5, 10. color solid 9.5; +#set terminal wxt size 950,1250 + +INDIR="build" +OUTDIR="build" + +set macros +set output OUTDIR."/plot-string-pbv.pdf" + +set multiplot layout 1, 2 ; + + +#set pointsize 2.0 +set grid +set key bottom right +#set key at screen 0.45, screen 0.8 +set xtics (1,2,5,10,20,50,100,200,500) +set logscale x +set logscale y 2 +set xlabel "String length passed, varying (mean)" +set ylabel "Time per pass (ns, mean), log_{2} scale" +set yrange [4:64] +set linetype 3 dashtype 2 +set linetype 4 dashtype 2 +plot INDIR."/plot-string-pbv-varcorp.dat" \ + i 0 using 1:2 title columnheader(1) with linespoints lt rgb "red" pt 3 ps 1 lw 1, \ + '' i 1 using 1:2 title columnheader(1) with linespoints lt rgb "blue" pt 6 ps 1 lw 1 + +set xlabel "String length passed, fixed" +set ylabel +plot INDIR."/plot-string-pbv-fixcorp.dat" \ + i 0 using 1:2 title columnheader(1) with linespoints lt rgb "red" pt 3 ps 1 lw 1, \ + '' i 1 using 1:2 title columnheader(1) with linespoints lt rgb "blue" pt 6 ps 1 lw 1 + +unset multiplot Index: doc/theses/mike_brooks_MMath/plots/string-peq-cppemu.py =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-peq-cppemu.py (revision 411aa65a1e6399ae182c88528a23be46ed2f8dd6) +++ doc/theses/mike_brooks_MMath/plots/string-peq-cppemu.py (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -12,7 +12,9 @@ import pandas as pd import numpy as np +import sys import os -infile = os.path.dirname(os.path.abspath(__file__)) + '/../benchmarks/string/result-append-pbv.csv' +sys.path.insert(0, os.path.dirname(__file__)) +from common import * prettyFieldNames = { @@ -23,58 +25,15 @@ } -timings = pd.read_csv( - infile, - names=['test', 'corpus', 'concatsPerReset', 'corpusItemCount', 'corpusMeanLenChars', 'concatDoneActualCount', 'execTimeActualSec'], - dtype={'test': str, - 'corpus': str, - 'concatsPerReset': 'Int64', # allows missing; https://stackoverflow.com/a/70626154 - 'corpusItemCount': np.int64, - 'corpusMeanLenChars': np.float64, - 'concatDoneActualCount': np.int64, - 'execTimeActualSec': np.float64}, - na_values=['xxx'], -) -# print(timings.head()) +timings = loadParseTimingData('result-append-pbv.csv') +# Filter operation=peq, corpus=100-*-1 -# project: parse executable and corpus names - -timings[['test-slug', - 'sut-platform', - 'operation', - 'sut-cfa-level', - 'sut-cfa-sharing', - 'op-alloc']] = timings['test'].str.strip().str.split('-', expand=True) -timings['sut'] = timings[['sut-platform', - 'sut-cfa-level', - 'sut-cfa-sharing', - 'op-alloc']].agg('-'.join, axis=1) - -timings[['corpus-basename', - 'corpus-ext']] = timings['corpus'].str.strip().str.split('.', expand=True) -timings[['corpus-slug', - 'corpus-nstrs', - 'corpus-meanlen', - 'corpus-runid']] = timings['corpus-basename'].str.strip().str.split('-', expand=True) -timings["corpus-nstrs"] = pd.to_numeric(timings["corpus-nstrs"]) -timings["corpus-meanlen"] = pd.to_numeric(timings["corpus-meanlen"]) -timings["corpus-runid"] = pd.to_numeric(timings["corpus-runid"]) - - -# project: calculate fact - -timings['op-duration-s'] = timings['execTimeActualSec'] / timings['concatDoneActualCount'] -timings['op-duration-ns'] = timings['op-duration-s'] * 1000 * 1000 * 1000 - - -# Filter operation=peq - -groupedOp = timings.groupby('operation') -tgtOpTimings = groupedOp.get_group('peq') - +timings = timings.groupby('operation').get_group('peq') +timings = timings.groupby('corpus-nstrs').get_group(100) +timings = timings.groupby('corpus-runid').get_group(1) # Emit in groups -groupedSut = tgtOpTimings.groupby('sut') +groupedSut = timings.groupby('sut') for sut, sgroup in groupedSut: Index: doc/theses/mike_brooks_MMath/plots/string-peq-sharing.d =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-peq-sharing.d (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-peq-sharing.d (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,2 @@ +plots/string-peq-sharing.gp.INPUTS: build/plot-string-peq-sharing.dat | build +plots/string-peq-sharing.py.INPUTS: benchmarks/string/result-append-pbv.csv Index: doc/theses/mike_brooks_MMath/plots/string-peq-sharing.gp =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-peq-sharing.gp (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-peq-sharing.gp (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,25 @@ +set terminal pdf color enhanced size 6.0in,3.0in font "Times,17" +#set terminal postscript portrait enhanced size 7.5, 10. color solid 9.5; +#set terminal wxt size 950,1250 + +INDIR="build" +OUTDIR="build" + +set macros +set output OUTDIR."/plot-string-peq-sharing.pdf" +#set pointsize 2.0 +set grid +set key top left +set xtics (1,2,5,10,20,50,100,200,500) +set logscale x +#set logscale y +set yrange [10:115] +set xlabel "String Length being appended (mean, geo. dist.), log scale" offset 2,0 +set ylabel "Time per append (ns, mean)" +set linetype 2 dashtype 2 +set linetype 4 dashtype 2 +plot INDIR."/plot-string-peq-sharing.dat" \ + i 0 using 1:2 title columnheader(1) with linespoints lt rgb "red" pt 2 ps 1 lw 1, \ + '' i 1 using 1:2 title columnheader(1) with linespoints lt rgb "red" pt 3 ps 1 lw 1, \ + '' i 2 using 1:2 title columnheader(1) with linespoints lt rgb "blue" pt 6 ps 1 lw 1, \ + '' i 3 using 1:2 title columnheader(1) with linespoints lt rgb "blue" pt 8 ps 1 lw 1 Index: doc/theses/mike_brooks_MMath/plots/string-peq-sharing.py =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-peq-sharing.py (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-peq-sharing.py (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,50 @@ +# Read thesis-append-pbv.csv +# Output for string-graph-peq-sharing.dat + +# Project details +# Filter operation=peq +# Split "series" goups of sut; only those in the "pretty" list +# Assert one row per string-length +# output: +# string-len op-duration +# in chunks, each headed by pertty(sut) + +import pandas as pd +import numpy as np +import os +import sys + +sys.path.insert(0, os.path.dirname(__file__)) +from common import * + +prettyFieldNames = { + "cfa-ll-share-fresh": "{/Helvetica=15 C{/Symbol \\42} +=} share fresh", + "cfa-ll-share-reuse": "{/Helvetica=15 C{/Symbol \\42} +=} share reuse", + "stl-na-na-fresh": "STL {/Helvetica=15 +=} fresh", + "stl-na-na-reuse": "STL {/Helvetica=15 +=} reuse", +} + +timings = loadParseTimingData('result-append-pbv.csv') + + +# Filter operation=peq, corpus=100-*-1 + +timings = timings.groupby('operation').get_group('peq') +timings = timings.groupby('corpus-nstrs').get_group(100) +timings = timings.groupby('corpus-runid').get_group(1) + + +# Emit in groups + +groupedSut = timings.groupby('sut') + +for sut, sgroup in groupedSut: + + if sut in prettyFieldNames: + + sgroup_sorted = sgroup.sort_values(by='corpus-meanlen') + + print('"{header}"'.format(header=prettyFieldNames[sut])) + text = sgroup_sorted[['corpus-meanlen', 'op-duration-ns']].to_csv(header=False, index=False, sep='\t') + print(text) + print() Index: doc/theses/mike_brooks_MMath/plots/string-pta-sharing.d =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-pta-sharing.d (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-pta-sharing.d (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,2 @@ +plots/string-pta-sharing.gp.INPUTS: build/plot-string-pta-sharing.dat | build +plots/string-pta-sharing.py.INPUTS: benchmarks/string/result-append-pbv.csv Index: doc/theses/mike_brooks_MMath/plots/string-pta-sharing.gp =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-pta-sharing.gp (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-pta-sharing.gp (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,24 @@ +set terminal pdf color enhanced size 6.0in,3.0in font "Times,17" +#set terminal postscript portrait enhanced size 7.5, 10. color solid 9.5; +#set terminal wxt size 950,1250 + +INDIR="build" +OUTDIR="build" + +set macros +set output OUTDIR."/plot-string-pta-sharing.pdf" +#set pointsize 2.0 +set grid +set key top left +set xtics (1,2,5,10,20,50,100,200,500) +set logscale x +set yrange [8:4096] +set logscale y 2 +set xlabel "String Length being appended (mean, geo. dist.), log scale" offset 2,0 +set ylabel "Time per append (ns, mean), log_{2} scale" +#show colornames +plot INDIR."/plot-string-pta-sharing.dat" \ + i 0 using 1:2 title columnheader(1) with linespoints lt rgb "red" pt 2 ps 1 lw 1, \ + '' i 1 using 1:2 title columnheader(1) with linespoints lt rgb "dark-green" pt 4 ps 1 lw 1, \ + '' i 2 using 1:2 title columnheader(1) with linespoints lt rgb "blue" pt 6 ps 1 lw 1, \ + '' i 3 using 1:2 title columnheader(1) with linespoints lt rgb "dark-green" pt 12 ps 1 lw 1 Index: doc/theses/mike_brooks_MMath/plots/string-pta-sharing.py =================================================================== --- doc/theses/mike_brooks_MMath/plots/string-pta-sharing.py (revision 241042418538c37dbb8067200a54e09c370946d2) +++ doc/theses/mike_brooks_MMath/plots/string-pta-sharing.py (revision 241042418538c37dbb8067200a54e09c370946d2) @@ -0,0 +1,57 @@ +# Read thesis-append-pbv.csv +# Output for string-graph-pta-sharing.dat + +# Project details +# Filter operation=peq +# Split "series" goups of sut; only those in the "pretty" list +# Assert one row per string-length +# output: +# string-len op-duration +# in chunks, each headed by pertty(sut) + +import pandas as pd +import numpy as np +import os +import sys + +sys.path.insert(0, os.path.dirname(__file__)) +from common import * + +prettyFieldNames = { + "peq": { + "cfa-ll-share-fresh": "{/Helvetica=15 C{/Symbol \\42} x += y} share fresh", + "stl-na-na-fresh": "STL {/Helvetica=15 x += y} fresh", + }, + "pta": { + "cfa-ll-share-fresh": "{/Helvetica=15 C{/Symbol \\42} x = x + y} share fresh", + "stl-na-na-fresh": "STL {/Helvetica=15 x = x + y} fresh", + } +} + +timings = loadParseTimingData('result-append-pbv.csv') + + +# Filter corpus=100-*-1 + +timings = timings.groupby('corpus-nstrs').get_group(100) +timings = timings.groupby('corpus-runid').get_group(1) + +# Emit in groups + +groupedSut = timings.groupby('sut') + +for sut, sgroup in groupedSut: + groupedOp = sgroup.groupby('operation') + for op,opPretty in prettyFieldNames.items(): + + if op in groupedOp.groups: + tgtOpTimings = groupedOp.get_group(op) + + if sut in opPretty: + + sgroup_sorted = tgtOpTimings.sort_values(by='corpus-meanlen') + + print('"{header}"'.format(header=opPretty[sut])) + text = sgroup_sorted[['corpus-meanlen', 'op-duration-ns']].to_csv(header=False, index=False, sep='\t') + print(text) + print()