diff --git a/plots/acf-pacf/implementations/python/pygal.py b/plots/acf-pacf/implementations/python/pygal.py index 5582e97374..7659c45a04 100644 --- a/plots/acf-pacf/implementations/python/pygal.py +++ b/plots/acf-pacf/implementations/python/pygal.py @@ -1,9 +1,18 @@ -""" pyplots.ai +""" anyplot.ai acf-pacf: Autocorrelation and Partial Autocorrelation (ACF/PACF) Plot -Library: pygal 3.1.0 | Python 3.14.3 -Quality: 85/100 | Created: 2026-03-14 +Library: pygal 3.1.0 | Python 3.13.13 +Quality: 85/100 | Updated: 2026-06-10 """ +import os +import sys + + +# Remove this file's directory from sys.path so `import pygal` resolves to the +# installed package, not this script (which shares the same name). +_this_dir = os.path.dirname(os.path.abspath(__file__)) +sys.path = [p for p in sys.path if os.path.abspath(p or ".") != _this_dir] + import io import xml.etree.ElementTree as ET @@ -15,7 +24,29 @@ from statsmodels.tsa.stattools import acf, pacf -# Data - Synthetic monthly airline-style passenger data with trend and seasonality +# Theme-adaptive tokens (Imprint palette) +THEME = os.getenv("ANYPLOT_THEME", "light") +PAGE_BG = "#FAF8F1" if THEME == "light" else "#1A1A17" +INK = "#1A1A17" if THEME == "light" else "#F0EFE8" +INK_MUTED = "#6B6A63" if THEME == "light" else "#A8A79F" + +# Imprint categorical palette — canonical order, theme-independent +IMPRINT_PALETTE = ("#009E73", "#C475FD", "#4467A3", "#BD8233", "#AE3030", "#2ABCCD", "#954477", "#99B314") +ANYPLOT_AMBER = "#DDCC77" # significance threshold lines (caution/warning role) + +# Semantic color roles for this chart +HIGHLIGHT_COLOR = IMPRINT_PALETTE[0] # #009E73 — significant bars (first series, always) +MUTED_BAR_COLOR = INK_MUTED # non-significant bars +CONF_LINE_COLOR = ANYPLOT_AMBER # 95% CI threshold lines +ZERO_LINE_COLOR = INK_MUTED # zero baseline + + +def hex_to_rgb(hex_str): + h = hex_str.lstrip("#") + return tuple(int(h[i : i + 2], 16) for i in (0, 2, 4)) + + +# Data — synthetic monthly airline-style passenger data with trend and seasonality np.random.seed(42) n_obs = 200 t = np.arange(n_obs) @@ -24,96 +55,94 @@ noise = np.random.normal(0, 2, n_obs) passengers = 100 + trend + seasonal + noise -# Compute ACF and PACF +# ACF/PACF computation n_lags = 36 acf_values = acf(passengers, nlags=n_lags, fft=True) pacf_values = pacf(passengers, nlags=n_lags, method="ywm") conf_bound = 1.96 / np.sqrt(n_obs) -# Color palette - refined for publication quality -highlight_color = "#306998" -muted_color = "#B8D4E8" -conf_line_color = "#C0392B" -bg_color = "#FAFAFA" -text_color = "#1A1A2E" -grid_color = "#E8ECF0" -zero_line_color = "#95A5A6" - +# Style — Imprint palette, theme-adaptive chrome; canvas = 3200×1800 (two 3200×900 panels) custom_style = Style( - background="white", - plot_background=bg_color, - foreground=text_color, - foreground_strong=text_color, - foreground_subtle=grid_color, - colors=(highlight_color,), - title_font_size=48, - label_font_size=30, - major_label_font_size=30, - legend_font_size=24, - value_font_size=18, + background=PAGE_BG, + plot_background=PAGE_BG, + foreground=INK, + foreground_strong=INK, + foreground_subtle=INK_MUTED, + colors=IMPRINT_PALETTE, + title_font_size=66, + label_font_size=56, + major_label_font_size=44, + legend_font_size=44, + value_font_size=36, + stroke_width=2.5, font_family="'Helvetica Neue', 'Helvetica', 'Arial', sans-serif", title_font_family="'Helvetica Neue', 'Helvetica', 'Arial', sans-serif", label_font_family="'Helvetica Neue', 'Helvetica', 'Arial', sans-serif", value_font_family="'Helvetica Neue', 'Helvetica', 'Arial', sans-serif", ) -# Y-axis ranges tailored to actual data with slight padding +# Y-axis ranges tailored to data acf_min, acf_max = -0.5, 1.1 pacf_min, pacf_max = -0.5, 1.0 -# Common chart config - high spacing for narrow stem-like bars +# Each panel: 3200×900; two panels stacked → 3200×1800 landscape common_config = { - "width": 4800, - "height": 1300, + "width": 3200, + "height": 900, "style": custom_style, "show_legend": False, "show_y_guides": True, "show_x_guides": False, - "margin": 25, - "margin_left": 120, + "margin": 20, + "margin_left": 130, "margin_right": 60, - "spacing": 100, + "spacing": 80, "truncate_label": -1, "print_values": False, "show_minor_x_labels": False, "x_labels_major_every": 4, "rounded_bars": 2, - "y_labels_major_count": 5, } -# ACF chart +# ACF chart (top panel) acf_chart = pygal.Bar( **common_config, x_title="", y_title="ACF", - title="acf-pacf · pygal · pyplots.ai", + title="acf-pacf · python · pygal · anyplot.ai", margin_bottom=10, margin_top=30, range=(acf_min, acf_max), ) acf_chart.x_labels = [str(i) for i in range(n_lags + 1)] +acf_chart.y_labels = [-0.4, -0.2, 0.0, 0.5, 1.0] acf_chart.add( "ACF", - [{"value": round(v, 4), "color": highlight_color if abs(v) > conf_bound else muted_color} for v in acf_values], + [{"value": round(v, 4), "color": HIGHLIGHT_COLOR if abs(v) > conf_bound else MUTED_BAR_COLOR} for v in acf_values], ) -# PACF chart +# PACF chart (bottom panel) pacf_chart = pygal.Bar( **common_config, x_title="Lag", y_title="PACF", title="", margin_bottom=70, margin_top=5, range=(pacf_min, pacf_max) ) pacf_chart.x_labels = [str(i) for i in range(1, n_lags + 1)] +pacf_chart.y_labels = [-0.4, -0.2, 0.0, 0.5, 1.0] pacf_chart.add( "PACF", - [{"value": round(v, 4), "color": highlight_color if abs(v) > conf_bound else muted_color} for v in pacf_values[1:]], + [ + {"value": round(v, 4), "color": HIGHLIGHT_COLOR if abs(v) > conf_bound else MUTED_BAR_COLOR} + for v in pacf_values[1:] + ], ) -# Render SVGs and inject confidence lines (pygal lacks native reference lines) +# SVG namespace registration ns = "http://www.w3.org/2000/svg" ET.register_namespace("", ns) ET.register_namespace("xlink", "http://www.w3.org/1999/xlink") -png_images = [] -for chart, y_min, y_max in [(acf_chart, acf_min, acf_max), (pacf_chart, pacf_min, pacf_max)]: + +def _inject_ci_lines(chart, y_min, y_max): + """Inject CI dashed lines and zero baseline into pygal's SVG (no native reference-line API).""" root = ET.fromstring(chart.render()) plot_group = next((g for g in root.iter(f"{{{ns}}}g") if g.get("class", "") == "plot"), None) if plot_group is not None: @@ -121,6 +150,7 @@ if bg_rect is not None: pw, ph = float(bg_rect.get("width")), float(bg_rect.get("height")) y_range = y_max - y_min + # 95% CI bounds (dashed amber lines) for cv in [conf_bound, -conf_bound]: ly = (y_max - cv) / y_range * ph line = ET.SubElement(plot_group, f"{{{ns}}}line") @@ -128,60 +158,67 @@ line.set("y1", f"{ly:.1f}") line.set("x2", str(pw)) line.set("y2", f"{ly:.1f}") - line.set("stroke", conf_line_color) + line.set("stroke", CONF_LINE_COLOR) line.set("stroke-width", "3") line.set("stroke-dasharray", "18,10") - line.set("opacity", "0.75") - # Zero baseline emphasis + line.set("opacity", "0.9") + # Zero baseline (muted solid line) zy = (y_max - 0) / y_range * ph zline = ET.SubElement(plot_group, f"{{{ns}}}line") zline.set("x1", "0") zline.set("y1", f"{zy:.1f}") zline.set("x2", str(pw)) zline.set("y2", f"{zy:.1f}") - zline.set("stroke", zero_line_color) + zline.set("stroke", ZERO_LINE_COLOR) zline.set("stroke-width", "2.5") zline.set("opacity", "0.6") + return root - png_images.append(cairosvg.svg2png(bytestring=ET.tostring(root), output_width=4800, output_height=1300)) -# Compose final image -acf_img = Image.open(io.BytesIO(png_images[0])) -pacf_img = Image.open(io.BytesIO(png_images[1])) -combined = Image.new("RGB", (4800, 2700), "white") -combined.paste(acf_img, (0, 50)) -combined.paste(pacf_img, (0, 1350)) - -# Subtle divider line between charts +# Render both panels +panels = [] +svg_strings = [] +for chart, y_min, y_max in [(acf_chart, acf_min, acf_max), (pacf_chart, pacf_min, pacf_max)]: + root = _inject_ci_lines(chart, y_min, y_max) + svg_bytes = ET.tostring(root) + svg_strings.append(ET.tostring(root, encoding="unicode")) + png_data = cairosvg.svg2png(bytestring=svg_bytes, output_width=3200, output_height=900) + img = Image.open(io.BytesIO(png_data)).convert("RGB") + img = img.resize((3200, 900), Image.LANCZOS) + panels.append(img) + +# Compose 3200×1800 landscape canvas +combined = Image.new("RGB", (3200, 1800), hex_to_rgb(PAGE_BG)) +combined.paste(panels[0], (0, 0)) +combined.paste(panels[1], (0, 900)) + +# Subtle divider between panels draw = ImageDraw.Draw(combined) -draw.line([(120, 1350), (4740, 1350)], fill="#DEE2E6", width=2) +draw.line([(130, 900), (3140, 900)], fill=hex_to_rgb(INK_MUTED), width=1) -# Confidence bound annotation - positioned in upper-right margin area +# CI annotation in upper-right margin try: font = ImageFont.truetype("/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf", 30) except OSError: font = ImageFont.load_default() -draw.text((4200, 68), f"95% CI: ±{conf_bound:.3f}", fill=conf_line_color, font=font) - -combined.save("plot.png", dpi=(300, 300)) +draw.text((2640, 24), f"95% CI: ±{conf_bound:.3f}", fill=CONF_LINE_COLOR, font=font) -# HTML version with interactive SVGs -acf_svg = acf_chart.render(is_unicode=True).replace('', "") -pacf_svg = pacf_chart.render(is_unicode=True).replace('', "") +combined.save(f"plot-{THEME}.png", dpi=(300, 300)) +# Interactive HTML output (pygal renders interactive SVG charts) html_content = ( "\n\n\n" - " acf-pacf · pygal · pyplots.ai\n" + " acf-pacf · python · pygal · anyplot.ai\n" " \n\n\n" "
\n" - f"
{acf_svg}
\n" - f"
{pacf_svg}
\n" + f"
{svg_strings[0]}
\n" + f"
{svg_strings[1]}
\n" "
\n\n" ) -with open("plot.html", "w") as f: +with open(f"plot-{THEME}.html", "w") as f: f.write(html_content) diff --git a/plots/acf-pacf/metadata/python/pygal.yaml b/plots/acf-pacf/metadata/python/pygal.yaml index 253adb0cb5..a20562c14a 100644 --- a/plots/acf-pacf/metadata/python/pygal.yaml +++ b/plots/acf-pacf/metadata/python/pygal.yaml @@ -1,44 +1,58 @@ library: pygal +language: python specification_id: acf-pacf created: '2026-03-14T22:18:53Z' -updated: '2026-03-14T22:56:02Z' -generated_by: claude-opus-4-5-20251101 -workflow_run: 23097477625 +updated: '2026-06-10T02:04:01Z' +generated_by: claude-sonnet +workflow_run: 27247413374 issue: 4663 -python_version: 3.14.3 +language_version: 3.13.13 library_version: 3.1.0 -preview_url: https://storage.googleapis.com/anyplot-images/plots/acf-pacf/pygal/plot.png -preview_html: https://storage.googleapis.com/anyplot-images/plots/acf-pacf/pygal/plot.html +preview_url_light: https://storage.googleapis.com/anyplot-images/plots/acf-pacf/python/pygal/plot-light.png +preview_url_dark: https://storage.googleapis.com/anyplot-images/plots/acf-pacf/python/pygal/plot-dark.png +preview_html_light: https://storage.googleapis.com/anyplot-images/plots/acf-pacf/python/pygal/plot-light.html +preview_html_dark: https://storage.googleapis.com/anyplot-images/plots/acf-pacf/python/pygal/plot-dark.html quality_score: 85 review: strengths: - - Excellent data storytelling through color differentiation of significant vs non-significant - lags - - Creative SVG post-processing to inject confidence interval lines (overcoming pygal - limitation) - - Perfect data quality with classic airline-style seasonal time series - - Clean dual-panel composition with shared x-axis concept - - HTML export preserves pygal native interactive tooltips + - 'Semantic color coding (significant=green #009E73, non-significant=gray) provides + immediate interpretive value without a legend' + - ANYPLOT_AMBER (#DDCC77) used correctly as a warning/caution anchor for 95% CI + threshold lines + - 'Both renders fully theme-adaptive: backgrounds, text, and gridlines correctly + flip between light (#FAF8F1) and dark (#1A1A17) themes' + - SVG injection via ElementTree is an elegant workaround for pygal's missing reference-line + API, enabling dashed CI lines + - Data (airline-style seasonal series, n=200) produces textbook ACF/PACF patterns + showcasing the plot type effectively + - 'Clean code: shared common_config dict reduces repetition; per-bar color dict + is idiomatic pygal usage' weaknesses: - - Bars are wider than ideal stem lines, though this is a pygal limitation - - Some wasted vertical space at bottom of canvas below Lag label - - SVG manipulation code adds complexity that slightly reduces code elegance - image_description: 'The plot displays two vertically stacked pygal bar charts on - a light gray (#FAFAFA) background. The top chart shows ACF values from lag 0 to - 36 with the title "acf-pacf · pygal · pyplots.ai" and y-axis labeled "ACF". The - bottom chart shows PACF values from lag 1 to 36 with y-axis labeled "PACF" and - x-axis labeled "Lag". Bars are colored in dark blue (#306998) when values exceed - the 95% confidence bounds, and light blue (#B8D4E8) when within bounds. Red dashed - horizontal lines at approximately ±0.139 indicate the 95% confidence interval. - A gray zero baseline is visible in both charts. A red annotation "95% CI: ±0.139" - appears in the upper-right corner. The ACF chart shows a clear sinusoidal decay - pattern with peaks at lags 0, 12, 24, 36 (period-12 seasonality). The PACF chart - shows a strong spike at lag 1 (~0.8), significant negative values at lags 2-4, - positive spikes around lags 7-10, and mostly non-significant values beyond lag - 12. A subtle gray divider separates the two charts.' + - Filled bars (pygal.Bar) instead of vertical stem lines — spec explicitly requires + 'vertical stem lines (not filled bars)'; consider very narrow bar widths or SVG-injected + vertical line segments to better approximate stems + - Non-uniform y-axis tick spacing (-0.4, -0.2, 0, 0.5, 1.0) — the 0→0.5 gap is 2.5× + the -0.2→0 gap, which can mislead scale perception; prefer uniform ticks like + [-0.4, -0.2, 0.0, 0.2, 0.4, 0.6, 0.8, 1.0] + - _inject_ci_lines helper function deviates from the KISS flat-structure ideal — + acceptable but noted + - 'Visual refinement could be stronger: panel spacing and whitespace utilization + inside each panel could be improved' + image_description: |- + Light render (plot-light.png): + Background: Warm off-white (~#FAF8F1) — correct theme surface + Chrome: Title "acf-pacf · python · pygal · anyplot.ai" in dark ink at top of ACF panel — readable. Y-axis labels "ACF" and "PACF" in dark ink on left — readable. X-axis "Lag" label at bottom in dark ink — readable. Tick labels (0, 4, 8, ... for ACF; 1, 5, 9, ... for PACF) in muted dark tone — readable. "95% CI: ±0.139" amber annotation in upper-right corner — readable. + Data: Bars colored brand green (#009E73) for significant correlations, muted gray for non-significant. ACF panel shows clear seasonal pattern (spikes at lags 12, 24, 36). PACF shows characteristic AR(1) spike at lag 1 and rapid decay. Amber dashed CI lines at ±0.139. Muted zero baseline line. First series uses #009E73 (correct). + Legibility verdict: PASS — all text and data elements clearly readable on light background. + + Dark render (plot-dark.png): + Background: Warm near-black (~#1A1A17) — correct dark theme surface + Chrome: Title in light cream color — clearly readable against dark background. Y-axis labels "ACF"/"PACF" in light text — readable. X-axis "Lag" label in light text — readable. Tick labels in light/muted-light text — readable. "95% CI: ±0.139" amber annotation visible. No dark-on-dark text failures detected. + Data: Bar colors identical to light render — brand green (#009E73) for significant, muted gray for non-significant. Data colors unchanged across themes (correct — only chrome flips). Amber dashed CI lines remain visible. + Legibility verdict: PASS — all text readable against dark background, no dark-on-dark failures. criteria_checklist: visual_quality: - score: 27 + score: 28 max: 30 items: - id: VQ-01 @@ -46,44 +60,51 @@ review: score: 7 max: 8 passed: true - comment: Font sizes explicitly set (title=48, labels=30, major_labels=30). - All text clearly readable. + comment: 'All text readable in both themes; minor: non-uniform y-axis tick + spacing (-0.4, -0.2, 0, 0.5, 1.0) subtly distorts scale perception' - id: VQ-02 name: No Overlap - score: 6 + score: 5 max: 6 passed: true - comment: No overlapping text. Major x-labels every 4 prevents crowding. + comment: No text overlaps; tightly packed bars in ACF lags 4-8 are dense but + no collision - id: VQ-03 name: Element Visibility - score: 5 + score: 6 max: 6 passed: true - comment: Bars clearly visible with good color contrast between significant - and non-significant values. + comment: Significant bars (green) and non-significant (gray) clearly distinguishable; + CI amber lines and zero baseline visible in both renders - id: VQ-04 name: Color Accessibility - score: 4 - max: 4 + score: 2 + max: 2 passed: true - comment: Blue/light blue scheme is colorblind-safe. Red confidence lines have - good contrast. + comment: Brand green / muted gray / amber are CVD-safe; no red-green as sole + signal - id: VQ-05 name: Layout & Canvas - score: 3 + score: 4 max: 4 passed: true - comment: Two stacked charts fill canvas reasonably well. Some wasted vertical - space below Lag label. + comment: Canvas gate passed (3200x1800); two-panel composition clean; no overflow + or clipping - id: VQ-06 name: Axis Labels & Title score: 2 max: 2 passed: true - comment: ACF, PACF, and Lag are descriptive and appropriate. Correlations - are unitless. + comment: ACF/PACF y-labels, Lag x-label, correct title format + - id: VQ-07 + name: Palette Compliance + score: 2 + max: 2 + passed: true + comment: 'First series #009E73; non-significant uses INK_MUTED anchor; CI + uses ANYPLOT_AMBER anchor; backgrounds correct for both themes' design_excellence: - score: 13 + score: 12 max: 20 items: - id: DE-01 @@ -91,22 +112,22 @@ review: score: 5 max: 8 passed: true - comment: Custom color palette with meaningful significance differentiation. - Above defaults but not publication-level. + comment: 'Above default: semantic color split, amber CI lines, corner annotation, + rounded bars show design intent' - id: DE-02 name: Visual Refinement - score: 4 + score: 3 max: 6 passed: true - comment: Subtle plot background, light grid lines, no x-guides, emphasized - zero baseline. + comment: Y-axis-only grid, minimal chrome, subtle panel divider; whitespace + and spine treatment could be stronger - id: DE-03 name: Data Storytelling score: 4 max: 6 passed: true - comment: Color differentiation between significant and non-significant lags - creates immediate visual hierarchy. + comment: Green/gray semantic split immediately directs eye to significant + lags; seasonal pattern and AR cutoff visually prominent spec_compliance: score: 14 max: 15 @@ -116,29 +137,28 @@ review: score: 4 max: 5 passed: true - comment: Correct dual-panel ACF/PACF layout. Bars instead of stems due to - pygal limitation. + comment: Correct ACF/PACF dual-panel concept; spec requires stem lines but + pygal.Bar produces filled bars — acceptable limitation - id: SC-02 name: Required Features score: 4 max: 4 passed: true - comment: 'All spec features present: ACF top, PACF bottom, 95% CI lines, correct - lag ranges.' + comment: ACF top/PACF bottom, 95% CI dashed lines, lag 0 in ACF, PACF from + lag 1, 36 lags all present - id: SC-03 name: Data Mapping score: 3 max: 3 passed: true - comment: X-axis shows lag numbers, Y-axis shows correlation values. Correctly - mapped. + comment: Lag on x-axis, correlation values on y-axis, all lags correctly displayed - id: SC-04 name: Title & Legend score: 3 max: 3 passed: true - comment: Title format correct. Legend appropriately hidden for single-series - charts. + comment: Title matches required format; legend omitted appropriately for semantically-colored + single-series data_quality: score: 15 max: 15 @@ -148,55 +168,58 @@ review: score: 6 max: 6 passed: true - comment: Shows seasonal pattern, AR component, positive/negative correlations, - significant and non-significant lags. + comment: Both ACF and PACF shown; CI lines; significant/non-significant coloring; + zero baseline; 36 lags - id: DQ-02 name: Realistic Context score: 5 max: 5 passed: true - comment: Airline-style passenger data with trend and seasonality. Classic - neutral time series example. + comment: Monthly airline-style passenger data with trend + 12-period seasonality; + neutral domain; 200 observations - id: DQ-03 name: Appropriate Scale score: 4 max: 4 passed: true - comment: 200 observations, realistic amplitudes, correlations in [-1, 1] range. + comment: Y-range [-0.5, 1.1] ACF / [-0.5, 1.0] PACF covers all values; 36 + lags appropriate for n=200 code_quality: score: 9 max: 10 items: - id: CQ-01 name: KISS Structure - score: 3 + score: 2 max: 3 passed: true - comment: Follows imports, data, plot, save structure. No functions or classes. + comment: 'Two helper functions: hex_to_rgb (small utility) and _inject_ci_lines + (SVG injection, justified by pygal API limitation)' - id: CQ-02 name: Reproducibility score: 2 max: 2 passed: true - comment: np.random.seed(42) set. + comment: np.random.seed(42) present - id: CQ-03 name: Clean Imports score: 2 max: 2 passed: true - comment: All imports used. + comment: All imports used; no dead imports - id: CQ-04 name: Code Elegance - score: 1 + score: 2 max: 2 passed: true - comment: SVG manipulation adds complexity but is justified by pygal limitations. + comment: Good use of common_config dict; per-bar color dict is idiomatic pygal; + sys.path workaround documented - id: CQ-05 name: Output & API score: 1 max: 1 passed: true - comment: Saves as plot.png and plot.html. Current API used. + comment: Saves plot-{THEME}.png and plot-{THEME}.html correctly library_mastery: score: 7 max: 10 @@ -206,26 +229,27 @@ review: score: 4 max: 5 passed: true - comment: Good use of pygal.Bar, Style class, per-value color dicts, x_labels_major_every, - rounded_bars. + comment: Good use of Style object, per-bar color dict, x_labels_major_every, + render() for SVG bytes - id: LM-02 name: Distinctive Features score: 3 max: 5 passed: true - comment: Per-value color customization, SVG rendering for post-processing, - HTML export with native interactivity. + comment: Per-bar color via dict syntax, HTML interactive output, SVG namespace + manipulation for CI injection verdict: APPROVED impl_tags: dependencies: - statsmodels - pillow techniques: + - subplots - annotations - html-export + - manual-ticks patterns: - data-generation dataprep: - time-series - styling: - - grid-styling + styling: []