Upgrade to Pro — share decks privately, control downloads, hide ads and more …

PyConJP2021に行ってきたログ.pdf

Intel0tw5727
August 02, 2022
11

 PyConJP2021に行ってきたログ.pdf

Intel0tw5727

August 02, 2022
Tweet

Transcript

  1.  !*OUFMUX • 1Z$PO+1͸τʔΫηογϣϯΛฉ͚͕ͩ͘શͯͰ͸ͳ͍ʂ • τʔΫηογϣϯҎ֎ʹ΋༷ʑͳίϯςϯπʹඈͼࠐΜͰΈΔ༐ؾ͕͋Ε͹ ָ͠Έ͸ഒ૿ʂ • 1Z$PO+1Λෆࣗ༝ͳָ͘͠Ή͜ͱ͕Ͱ͖Δͷ͸ɺӡӦελοϑਓਓͷؤ ுΓͷ͓͔͛ͩͬͨΓ͢ΔͷͰɺ࠷େݶͷײँΛ👏

    • ࢀՃऀͱͯ͠෺଍Γͳ͘ͳͬͨΒɺ࣍͸εϙϯαʔ΍ελοϑ΍ଞͷܗͰࢀ Ճͯ͠Έ·͠ΐ͏ʂ 5-%3 ͜ͷൃදͰ఻͍͑ͨ͜ͱ 1Z$PO+1ͷָ͠Έํ͸ઍࠩສผɻࣗ෼ͳΓͷָ͠ΈํΛ୳͠ʹདྷ ೥΋ͥͻࢀՃͯ͠ΈΑ͏ʂ
  2.  !*OUFMUX • ؆୯ͳΠϕϯτ঺հ • τʔΫηογϣϯ΁ͷײ૝ • τʔΫηογϣϯҎ֎ͷίϯςϯπʹ͍ͭͯ • ӡӦܦݧऀ໨ઢͰݟͨ1Z$PO+1

    • ϓϥνφεϙϯαʔʹͳΔ࣮ͬͯࡍͲ͏ͳͷʁ ຊ೔࿩͢͜ͱ ओʹ1Z$PO+1ͷ͍ΖΜͳίϯςϯπΛࢀՃऀ໨ઢͰ঺հ͍ͯ͘͠ ͜ͱ͕ϝΠϯςʔϚͰ͢ʂ
  3.  !*OUFMUX • ͪΎΒσʔλגࣜձࣾ ݩؾসإσʔλΞφϦετ • )/͍ΜͯΔ Ͱ׆ಈ͍ͯ͠·͢ ʢ4/4!JOUFMUX •

    1Z$PO ,ZVTIVJO0LJOBXB࠲௕ 1Z$PO ,ZVTIV໾һɺ1Z%BUB0LJOBXB ΦʔΨφΠβʔ • ޷͖ͳϥΠϒϥϦ͸ʮURENʯ ࣗݾ঺հ ଟ࿨ా ਅޛ r 5BXBEB 4IJOHP
  4.  !*OUFMUX • 1Z$PO+1 ೔ຊશࠃ͔Β1ZUIPOJTUB͕ू·Δ1Z$PO • 1Z$PO NJOJ ֤஍ҬͰ։࠵͞ΕΔ1Z$PO •

    )JSPTIJNB • 4IJ[VPLB • 0TBLB • 1Z$PO ,ZVTIV ۝भ஍۠Ͱ։࠵͞ΕΔ1Z$PO • %KBOHP$POHSFTT ʮ%KBOHPʯϑϨʔϜϫʔΫʹؔΘΔશͯͷਓͷΧϯϑΝϨϯε Πϕϯτ঺հ 1Z$PO+1ͱ͸೔ຊͰ։࠵͞ΕΔ1ZUIPOϢʔβʔͷͨΊͷΧϯ ϑΝϨϯεͰ৘ใަ׵΍ަྲྀΛ͢ΔΠϕϯτͰ͢ʂ
  5.  !*OUFMUX • ୩߹ ኍل͞Μ ʢকعͱ1ZUIPOͷૉఢͳग़ձ͍ʣ • ϓϩع࢜ º ΤϯδχΞ

    ͱ͍͏ҟ৭ͷ૊Έ߹Θͤͳ͜ͷํ͸ɺϓϩع࢜ͱ͸͔͚Β΋ײ ͤ͡͞ͳ͍ΤϯδχΞ৭ͷೱ͍ํͰɺʮ1ZUIPOº কعͰԿ͕ղܾͰ͖ΔͩΖ͏ʁʯͱ ৗʹνϟϨϯδ͢Δૉ੖Β͍͠ํͰͨ͠ɻ τʔΫηογϣϯ΁ͷײ૝ ೔໨͸ϓϩع࢜ͷ୩߹ኍل͞Μͷ,FZOPUF͔Β࢝·Γɺσʔλ ෼ੳؔ࿈ͷηογϣϯΛத৺ʹฉ͍ͨ೔
  6.  !*OUFMUX • 'VKJOF 4IJHFOPCV͞Μ ʢTDJLJUMFBSOͷ৽ػೳΛ঺հ͠·͢ʣ • σʔλ෼ੳʹ͓͍ͯଉΛٵ͏Α͏ʹ࢖͍ͬͯΔ4DJLJUMFBSOϥΠϒϥϦʹ͍ͭͯɺ௚ۙ௥ Ճ͞ΕͨػೳΛ۩ମྫͱͱ΋ʹ؆ܿʹ঺հ͍͚ͨͩͨൃදͰɺͳʹ͔࣋ͪؼΔ͜ͱ͕Ͱ ͖ͨࢀՃऀ΋গͳ͘ͳ͍͸ͣɻ

    τʔΫηογϣϯ΁ͷײ૝ ೔໨͸ϓϩع࢜ͷ୩߹ኍل͞Μͷ,FZOPUF͔Β࢝·Γɺσʔλ ෼ੳؔ࿈ͷηογϣϯΛத৺ʹฉ͍ͨ೔ )BMWJOH(SJE4FBSDI$7 param_grid = { "max_depth": [3, None], "min_samples_split": [5, 10] } search = HalvingGridSearchCV( clf, param_grid, resource='n_estimators', max_resources=10, random_state=0 ).fit(X, y) $PMVNO5SBOTGPSNFS featuring = ColumnTransformer([ ('std', StandardScaler(), range(10)), ('label', LabelEncoderM(), (10, )) ])
  7.  !*OUFMUX • ,PZBNB5FUTVP͞Μ 7JTVBMJ[F%TDJFOUJGJDEBUBJOB1ZUIPOJDXBZMJLFNBUQMPUMJC • %ը૾ͷॲཧ΍ՄࢹԽ͕Ͱ͖Δ7JTVBMJ[BUJPO5PPM,JU 75, Λɺ/VNQZ഑ྻͰѻ͑Δ Α͏ʹͨ͠1Z7JTUBͱ͍͏ϥΠϒϥϦͷ঺հͰɺ1ZUIPO্Ͱ%ը૾Λ͜Ͷ͜Ͷ͢Δʹ

    ͸ॿ͔Γͦ͏ͳϥΠϒϥϦͰͨ͠ɻ τʔΫηογϣϯ΁ͷײ૝ ೔໨͸ϓϩع࢜ͷ୩߹ኍل͞Μͷ,FZOPUF͔Β࢝·Γɺσʔλ ෼ੳؔ࿈ͷηογϣϯΛத৺ʹฉ͍ͨ೔ 45 Conclusion We introduce the PyVista’s • Pythonic interface to VTK’s Python bindings • Filtering/plotting tools built for interactivity • Direct access to common VTK filters • Intuitive plotting routines with matplotlib similar syntax In this presentation, you can learn more about how PyVista wraps di↵erent VTK mesh types and how you can leverage powerful 3D plotting and mesh analysis tools. Highlights of the API include: Pythonic interface to VTK’s Python bindings Filtering/plotting tools built for interactivity (see Widgets) Direct access to common VTK filters (see Filters) Intuitive plotting routines with matplotlib similar syntax (see Plotting)
  8.  !*OUFMUX • ϒϥϯτ ϒʔΧʔࢯ͞Μ "1FSGFDUNBUDI • 1ZUIPOͷύλʔϯϚονΛ4DBMB΍3VTUͷΑ͏ͳʮNBUDIr DBTFʯߏจͰ࣮૷ͨ͠࿩ Ͱɺ࣮ࡍͷ࢖༻ྫΛࣔ͠ͳ͕ΒैདྷͷʮJGr

    FMTFʯߏจͱൺֱͯ͠εϚʔτʹΫʔϧʹ ࠇຐज़νοΫʹʁ͔͚ΔΑ͏ͳͱ͜ΖΛ঺հͯ͘͠Ε·ͨ͠ɻ͔ΒͷػೳͳͷͰૣ ͘࢖ͬͯΈ͍ͨͰ͢Ͷɻ τʔΫηογϣϯ΁ͷײ૝ ೔໨͸ύλʔϯϚονϯά։ൃ࿩͔Β࢝·Γɺσʔλ෼ੳҎ֎ͷ ڵຯ͋ΔηογϣϯΛத৺ʹฉ͍ͨ೔ Syntax The Design // Rust fn f(n: u64) -> u64 { match n { 0 | 1 => 1, _ => n * f(n - 1), } } // Scala def f(n: Int): Int = n match { case 0 | 1 => 1 case _ => n * f(n - 1) } # Python def f(n: int) -> int: match n: case 0 | 1: return 1 case _: return n * f(n - 1)
  9.  !*OUFMUX • TIJOZPSLF͞Μ ࣮ફ4USFBNMJU 'MBTL "*ϓϩδΣΫτͷϓϩτλΠϐϯά͔Βຊ൪ӡ༻·ͰΛ͍͍ײ͡ʹ͢Δ 1ZUIPOJDͳ΍Γ͔ͨ • 4USFBNMJUͱ'MBTLΛ࢖ͬͯɺಈ͘ΞϓϦέʔγϣϯΛαΫαΫ࡞͍͖ͬͯͳ͕Βɺվળ

    αΠΫϧΛΨϯΨϯճ͍ͯ͜͠͏ͱ͍͏঺հͰɺϓϩδΣΫτͱͯ͠΋ෆ࣮֬ੑͷଟ͍ ͱ͜ΖʹΞδϟΠϧͳ։ൃΛ౰͍͖ͯͯ·͠ΐ͏ͱ͍͏঺հͰͨ͠ɻ τʔΫηογϣϯ΁ͷײ૝ ೔໨͸ύλʔϯϚονϯά։ൃ࿩͔Β࢝·Γɺσʔλ෼ੳҎ֎ͷ ڵຯ͋ΔηογϣϯΛத৺ʹฉ͍ͨ೔ ʮ͍͍͔Μ͡ʹ͢ΔPythonicͳ΍Γ͔ͨʯ is ʮαΫοͱಈ͘΋ͷΛ࡞ͬͯ΍ͬͯ͜(ʯ ΨϯΨϯͱσϦόϦʔ͍ͯ͜͠͏ʂ AIϓϩδΣΫτʹඞཁͳελϯε • ෆ࣮֬ͳϓϩδΣΫτ͸Agileͳ΍ΓํͰղܾ͍ͯ͘͠ • σʔλαΠΤϯςΟετͱΤϯδχΞ͸ҧ͏ੜଶܥͷੜ͖෺ νʔϜϫʔΫΛେ੾ʹʂ ͜ͷൃදͰҰ؏͍ͯ͠Δߟ͑ํɾେ੾ʹͯ͠ΔࣄͷએݴͰ͢ ʢҟ࿦͸ೝΊΔʣ
  10.  !*OUFMUX • ͻΖ͞͡͞Μ ֆΛಡΉٕज़ 1ZUIPOʹΑΔΠϥετղੳ • Πϥετͷߏਤ΍ߏ੒ʹ͍ͭͯɺॳ৺ऀͰ΋Θ͔ΔΑ͏ͳஸೡͳઆ໌͔Βɺ࣮ࡍʹΠϥ ετղੳʹ͸ͲͷΑ͏ʹϥΠϒϥϦΛ͔͖͔ͭͬͯͨΛ঺հͯ͘͠ΕͨൃදͰͨ͠ɻ Πϥετ͔Θ͍͍ʂʣ

    τʔΫηογϣϯ΁ͷײ૝ ೔໨͸ύλʔϯϚονϯά։ൃ࿩͔Β࢝·Γɺσʔλ෼ੳҎ֎ͷ ڵຯ͋ΔηογϣϯΛத৺ʹฉ͍ͨ೔ Ϋϥε෼ྨث͕ͲͷྖҬΛ΋ͱʹը૾Λ෼ྨ͢Δ͔ΛՄࢹԽ͢Δ$MBTT"DUJWBUJPO.BQʹͯ࠶ݱɻ PythonͰয఺Λݕग़͢ΔᶄɿҰ෦ղઆ য఺ʮݟ׳ΕͨϞϊʯͷݕग़ʢUGLFSBTWJTʹͯ(SBE$". Λར༻ʣ # import libraries (ུ) # prepare model & input data model = Model(weights='imagenet', include_top=True) image = load_img(img_name, target_size=(224, 224)) X = preprocess_input(np.array(image)) # set loss & modifier to replace a softmax function def loss(output): return (output[0][cls_index]) def model_modifier(m): m.layers[-1].activation = tf.keras.activations.linear return m # generate heatmap with GradCAM++ gradcam = GradcamPlusPlus(model, model_modifier=model_modifier, clone=False) cam = gradcam(loss, img, penultimate_layer=-1) cam = normalize(cam) heatmap = np.uint8(cm.jet(cam[0])[..., :3] * 255) LFJTFOUGLFSBTWJTc(JUIVC
  11.  !*OUFMUX • ͕࣌ؒͳ͘શ෦ΛճΔ͜ͱ͸೉͔ͬ͠ ͨͷͰ͕͢ɺଟ͘ͷاۀ୲౰ͷํͱ ʮίϩφӔͰมΘͬͨಇ͖ํʯ΍ʮا ۀͷ࣋ͭϢχʔΫͳจԽʯͳͲͰ੝Γ ্͕Δ͜ͱ͕Ͱ͖·ͨ͠ɻ • ଟࠃ੶ͳࣾһ͕ॴଐ͢Δ)&//(&͞Μɺ

    ࠓճͷ(VJEF͔ΒͷϝοηʔδΛ͍ͨ ͩ͘·Ͱͷཪ࿩Λͯ͘͠Εͨ.JDSPTPGU ͞Μɺຖ೥ࢀՃܕͷήʔϜΛ४උͯ͠ ͘ΕΔ+9௨৴ࣾ͞Μɺ௕໺ͷ౔஍Ͱ 1ZUIPOίϛϡχςΟΛ׆ൃʹ͍ͯ͠Δ ೔ຊγεςϜٕݚ͞ΜͳͲͳͲɻ τʔΫηογϣϯҎ֎ͷ࣌ؒ͸ԿΛʁ اۀϒʔεΛճͬͯΈͨΓɺηογϣϯεϐʔΧʔʹ࣭໰ͨ͠Γɺ ࣗࣾͷεϙϯαʔϒʔεͰײ૝ΛޠͬͨΓ͍ͯ͠·ͨ͠ɻ اۀϒʔε ϒʔεπΞʔ ௨ৗ࣌ͷདྷ৔ ϒʔεπΞʔ࣌ͷདྷ৔ ׬શʹఢऻɾɾɾXX
  12.  !*OUFMUX • "TL'PS4QFBLFS • τʔΫηογϣϯऴྃޙͷ໿෼͸ɺൃදऀʹ௚઀࣭໰ͨ͠Γײ૝Λड़΂Δ͜ͱ͕Ͱ͖ Δ࣌ؒͰͨ͠ɻ • ൃද࣌ؒதʹ࣭໰Ͱ͖Ε͹ྑ͍ͷͰ͕͢ɺͳ͔ͳ͔͕࣌ؒ୹͍ͨΊ࣭໰Λ੔ཧ͍ͯ͠Δ ؒʹऴΘͬͯ͠·͏ࢲʹ͸͋Γ͕͍ͨ࣌ؒͰͨ͠ɻ

    τʔΫηογϣϯҎ֎ͷ࣌ؒ͸ԿΛʁ اۀϒʔεΛճͬͯΈͨΓɺηογϣϯεϐʔΧʔʹ࣭໰ͨ͠Γɺ ࣗࣾͷεϙϯαʔϒʔεͰײ૝ΛޠͬͨΓ͍ͯ͠·ͨ͠ɻ ͳΜͱ͔ൃදதʹ಄ͷ੔ཧ͕Ͱ͖࣭ͨ໰ ͜ͷ࣭໰Ҏ֎͸"TL'PS4QFBLFSͰ΍ͬͨΓ͠ ͯ·ͨ͠
  13.  !*OUFMUX ຖ೥Πϕϯτதʹɺࡢ೥౓·Ͱͷ׆ಈใࠂɺ ࠓޙͷܭը1Z$PO+1 ʹ͍ͭͯͳͲɺ ͜Ε͔Βͷ1Z$PO+1ʹ͍ͭͯɺཧࣄͱࢀՃ ऀΛަ͑ͯձٞΛ͍ͯ͠·͢ɻ Πϕϯτ౰೔ࢀՃͰ͖ͳ͔ͬͨίϯςϯπ ͜ͷଞʹ΋ʮϥϯνλΠϜηογϣϯʯ΍ʮ1Z$PO+1 ެ։ӡӦ

    ձٞʯͳͲ͕͋Γ·ͨ͠ɻ ϥϯνλΠϜηογϣϯ ެ։ӡӦձٞ ࠓճϓϥνφεϙϯαʔʹཱީิͨ͠ࡾࣾ୅ දऀʹΑΔύωϧσΟεΧογϣϯΛ࣮ࢪ͠ ͯɺίϩφӔͰͷಇ͖ํ΍1ZUIPOͱͷग़ձ ͍ʹ͍ͭͯ࿩͍͍ͯͨͩͯ͠·ͨ͠ɻ
  14.  !*OUFMUX • %JTDPSE ͱ ;PPNΛ༻͍ͨϋΠϒϦουͳମ੍ • ;PPN͸τʔΫηογϣϯผʹ෼͚ΒΕͨ෦԰ • %JTDPSE͸ࢀՃऀؒͷίϛϡχέʔγϣϯͱͯ͠੾Γ෼͚ΒΕͨମ੍

    • ΦϑϥΠϯͰ͔͠Ͱ͖ͳ͍ʮମݧʯΛɺΦϯϥΠϯͰ୅ସͰ͖Δʮମݧʯ΁ ੾Γସ͑ͨੵۃతͳࢪࡦ • ௌऺͷϦΞΫγϣϯΛνϟϯωϧͷεϨουͱͯ͠ྲྀ͢ํࣜ΁ • ϒʔεπΞʔʹΑ֤ͬͯεϙϯαʔͷձࣾ΁ͷ઀৮ػձ • ࠙਌ձ͸QJ[[BIVU͞Μͱ࿈ܞͯ͠ɺࢀՃऀ΁ͷϑʔυσϦόϦʔΦϯϥΠϯҿΈ΁ • ౰೔ͷࢀՃऀ࣭໰ɾཁ๬ରԠ • 1Z$PO+1Λָ͠ΜͰ΋Β͏ͨΊͷͰ͖ΔΞΠσΟΞΛܗʹ͢ΔӡӦύϫʔ ӡӦܦݧऀ໨ઢͰݟͨ1Z$PO+1 ഑৴Λ࢝Ίͱͯ͠ɺΦϯϥΠϯ্ͰͷࢀՃऀ༠ಋɺ֤Πϕϯτ঺ հɺӡӦ΁ͷ࣭໰ɾཁ๬ରԠͳͲɺεϜʔζͳΠϕϯτͰͨ͠
  15.  !*OUFMUX ü഑৴ଆΛ;PPNʹ౷Ұ͢Δ͜ͱͰࢀՃऀ΋ ඞཁͳ෦԰ʹҠಈ͢Δ·Ͱ͕ϦϯΫͭͰ Ͱ͖͍ͯ·ͨ͠ɻ üԻ੠ΛϚϧννϟϯωϧʹͯ͠,FZOPUF຋ ༁Λ͍ͯͨ͠ͷ͸ͱͯ΋͍͍ΞΠσΟΞͩ ͱࢥ͍·ͨ͠ɻ ӡӦܦݧऀ໨ઢͰݟͨ1Z$PO+1 ֤ձ৔ͱϦϯΫͨ͠%JTDPSEͷνϟϯωϧͰ͸ɺτʔΫηογϣ

    ϯ΁ͷײ૝͕ྲྀΕ͍ͯͯฉ͚ͩ͘Ͱͳ͘ࢀՃͰ͖Δܗʹ ;PPN %JTDPSE ü֤τʔΫηογϣϯͰͷײ૝͕֤νϟϯω ϧΛྲྀΕΔ͜ͱͰɺ:PVUVCF -JWFͷΑ͏ ͳࢀՃܕͷମݧ͕ͱͯ΋Α͔ͬͨͰ͢ɻ ü࣭໰Λνϟϯωϧ͔Β࢘ձ͕र্͍͛Δํ ࣜʹ͢Δ͜ͱͰɺ࣭໰΁ͷෑډΛ௿ͯ͘͠ ΑΓଟ͘ͷ࣭໰͕ू·ͬͨͷͰ͸ͳ͍͔ͱ ࢥͬͨΓ͠·ͨ͠ɻ
  16.  !*OUFMUX • ڈ೥͕;PPN੾Γ෼͚ͩͬͨͷʹର ͯ͠ɺ%JTDPSEͷΑ͏ͳ֤ϒʔεͷ ਓ਺ঢ়گ͕೺ѲͰ͖Δܗʹͳ͓ͬͨ ͔͛ͰɺؾܰʹೖΓ΍͔ͬͨ͢ͱࢥ ͍·ͨ͠ɻ • ελϯϓϥϦʔ΍ϒʔεπΞʔΛઃ

    ͚Δ͜ͱͰɺϒʔεʹཱͪدΔ͖ͬ ͔͚Λଟ͘࡞ͬͯ͘Εͨͷ͸ɺ͔ͳ Γ͋Γ͕͔ͨͬͨͰ͢ɻ ӡӦܦݧऀ໨ઢͰݟͨ1Z$PO+1 ϒʔεπΞʔ΍֤ϒʔεͷਓ਺ঢ়گ͕ࢀՃऀ͔Βݟ͑Δ͜ͱ͔Β ଍Λӡͼ΍͍͢ҹ৅
  17.  !*OUFMUX • νέοτߪೖஈ֊Ͱ഑ୡઌࢦఆͨ͠Γɺର৅۠Ҭ֎ͷ஍Ҭ͸ผҊ΋ݕ౼ͨ͠ Γ͢ΔͳͲɺ͔ͳΓؤுͬͨͱࢥ͍·͢ɻ • ࣮ࡍࢀՃऀͷ͏ͪɺ͏·͘ಧ͔ͳ͔ͬͨέʔε͕ͲΕ͘Β͍͋Δͷ͔ͳʁͱ ࢥͬͨͷͰɺ΋͠ৼΓฦΓͱ͔Ͱ͜ͷ࿩୊͕ग़͖ͯͨΒɺӡӦଆ͕ͲΕ͘Β ͍ෛ୲ͩͬͨͷ͔͸ฉ͍ͯΈ͍ͨͱ͜Ζɻ •

    ͪͳΈʹࢲ͸ର৅۠Ҭ֎ͰσϦόϦʔ͕೉͘͠அ೦͍ͯͨ͠ͷͰ͕͢ɺϐβ ςϩͷμϝʔδ͸େ͖͔ͬͨͷͰɺ݁ہυϛϊɾϐβ🍕པΈ·ͨ͠ɻඒຯ͠ ͔ͬͨͰ͢ɻ ԭೄʹ͸ϐβϋοτ͕ళฮ͋Δ΋ͷͷɺ͢΂ͯσϦόϦʔ͕ͳ͍ͨΊஅ ೦ɾɾɾϐβϋοτ͞ΜͲ͏ͧΑΖ͓͘͠ئ͍͠·͢🙇 ӡӦܦݧऀ໨ઢͰݟͨ1Z$PO+1 ࠙਌ձ͸ϐβϋοτ͞Μͱ࿈ܞͯ͠ɺࢀՃऀ΁ͷϑʔυσϦό ϦʔΦϯϥΠϯҿΈ΁
  18.  !*OUFMUX ࣾ಺͸ϓϥνφܾఆͷ࣌ʹେ੝Γ͕͋Γ ·͔͞ϓϥνφεϙϯαʔ௨Δͱ͸ࢥ͓ͬͯΒͣɺ֖Λ։͚ͯΈ ͨΒͲ͏΍Β໊ͩͨΔاۀʹڬ·Ε͍ͯͨ໛༷ ೥Ҏ্લ͔Β1Z$PO "1"$΍ 1Z$PO +1΁εϙϯαʔΛଓ͚͍ͯ Δݹࢀεϙϯαʔɻࢲ͕ॳΊͯࢀՃ

    ͨ͠1Z$PO͕೥ͩͬͨͷͰ͢ ͕ɺ೥ʹձ໊ࣾশ͕มΘͬͨ Β͘͠ɺͦΕΛϒʔεʹ༡ͼʹߦͬ ͨͱ͖ʹ஌ͬͯͼͬ͘Γɻ )&//(&גࣜձࣾ͞Μ ݴΘͣͱ΋஌Εͨ༗໊ͳاۀͰɺ )&//(&͞Μಉ༷ʹ1Z$PO +1΁ε ϙϯαʔΛଓ͚͍ͯΔݹࢀεϙϯ αʔɻࠓճͷ(VJEP͞Μ͔Βͷϝο ηʔδΛ໯͏·ͰʹճҎ্ϝʔϧ ͨ͠ͱ͔ͳΜͱ͔ɻ ೔ຊϚΠΫϩιϑτגࣜձࣾ͞Μ ͪΎΒσʔλגࣜձࣾ ձࣾ঺հ$.ͰಥવԭೄํݴΛ࿩͠ ࢝Ίɺԭೄݝ֎ࢀՃऀʹδϟϒΛ ଧ͍ͬͯ͘ελΠϧͷاۀɻ
  19.  !*OUFMUX • 1Z$PO+1͸τʔΫηογϣϯΛฉ͚͕ͩ͘શͯͰ͸ͳ͍ʂ • τʔΫηογϣϯҎ֎ʹ΋༷ʑͳίϯςϯπʹඈͼࠐΜͰΈΔ༐ؾ͕͋Ε͹ ָ͠Έ͸ഒ૿ʂ • 1Z$PO+1Λෆࣗ༝ͳָ͘͠Ή͜ͱ͕Ͱ͖Δͷ͸ɺӡӦελοϑਓਓͷؤ ுΓͷ͓͔͛ͩͬͨΓ͢ΔͷͰɺ࠷େݶͷײँΛ👏

    • ࢀՃऀͱͯ͠෺଍Γͳ͘ͳͬͨΒɺ࣍͸εϙϯαʔ΍ελοϑ΍ଞͷܗͰࢀ Ճͯ͠Έ·͠ΐ͏ʂ ·ͱΊʢ࠶ܝʣ 1Z$PO+1ͷָ͠Έํ͸ઍࠩສผɻࣗ෼ͳΓͷָ͠ΈํΛ୳͠ʹདྷ ೥΋ͥͻࢀՃͯ͠ΈΑ͏ʂ
  20.  !*OUFMUX • ೔࣌೥݄೔ ౔ • ձ৔۽ຊ৓ϗʔϧ • Ωʔϊʔτਗ਼ਫ઒ و೭͞Μ

    • $G1ΛઈࢍืूதͰ͢ʂ • 1Z$PO+1ͰൃදͰ͖ͳ͔ͬͨํ • ۝भʹΏ͔Γͷ͋Δํ • ۝भʹΏ͔Γͷͳ͍ํ • ͥͻͥͻ͓଴͍ͪͯ͠·͢ʂ 1Z$PO ,ZVTIVJO Λ։࠵͠·͢ʂ ࠓ೥౓͸1Z$PO ,ZVTIVΛ۽ຊ৓ϗʔϧʹ݄ͯ೔ ౔ ʹ։࠵ ͍ͨ͠·͢ʂ