visualizations
Guides missingno matrix, bar, heatmap, and dendrogram missing-data visualization workflows, configuration, interpretation, and plotting troubleshooting.
- License 许可
- MIT
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- visualizations
- license
- MIT
- metadata
- { "disco-role": "operating" }
- description
- Guides missingno matrix, bar, heatmap, and dendrogram missing-data visualization workflows, configuration, interpretation, and plotting troubleshooting.
- disable-model-invocation
- true
02 资源文件 3
- 参考文档references/api-reference.md4.3 KB
- 参考文档references/troubleshooting.md3.4 KB
- 参考文档references/workflows.md4.2 KB
03 SKILL.md 正文
missingno Visualizations
Use this sub-skill when a task asks how to make, configure, interpret, save, or
troubleshoot missingno plots for pandas DataFrames.
Choose the plot
| Need | Use | Why |
|---|---|---|
| See row-by-row and column-by-column missingness patterns | msno.matrix(df, ...) |
Dense nullity matrix with optional sparkline for row completeness. |
| Compare per-column completeness counts/ratios | msno.bar(df, ...) |
Simpler bar view for variable completeness, with optional log scale. |
| Find pairwise nullity correlation between variables | msno.heatmap(df, ...) |
Correlates missing/present masks and annotates relationships. |
| Group variables by similar missingness patterns | msno.dendrogram(df, ...) |
Uses SciPy hierarchical clustering of nullity masks. |
For column filtering or completeness sorting before plotting, read ../nullity-utilities/SKILL.md.
Fast workflow
import matplotlib.pyplot as plt
import missingno as msno
# df is a pandas.DataFrame.
ax = msno.matrix(df, sparkline=False)
ax.figure.savefig("missingness-matrix.png", bbox_inches="tight")
plt.close(ax.figure)
In headless environments, set MPLBACKEND=Agg before running the script. For a
safe end-to-end package check, run the root helper:
../../scripts/missingno_smoke_check.py.
References
- Read references/workflows.md for plot-selection recipes, interpretation notes, saving figures, large-column handling, and time-index examples.
- Read references/api-reference.md for verified function signatures, parameter behavior, return values, and source-backed implementation notes.
- Read references/troubleshooting.md for
headless matplotlib, label overlap,
freqerrors, dropped heatmap columns, staleinline/geoplotmentions, and dependency issues.
Decision points
- If a user asks for a single visual overview, start with
matrixfor small to medium column counts andbarwhen only completeness percentages are needed. - If a user asks whether missingness in two columns is related, use
heatmap; explain that columns that are always full or always empty are removed before correlation. - If a user asks about broader missingness groups or many variables, use
dendrogram; consider filtering to relevant columns first. - If a plot is unreadable because there are many variables, reduce labels,
change orientation where supported, or route to
../nullity-utilities/SKILL.md for
filter='bottom',n=..., andp=...preparation. - If the task uses a time index with
matrix(freq=...), confirm the DataFrame index is aPeriodIndexorDatetimeIndexand that the requested frequency produces ticks found in the index range.
Boundaries
- This sub-skill does not cover generic seaborn/matplotlib styling except where needed to save/customize returned axes.
- This sub-skill does not claim
geoplotorinlinesupport for this snapshot. - This sub-skill does not require network sample data; use small synthetic frames for examples and checks.
04 同仓库更多 Skills
passnet-triton-opt
Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.
passnet-solve
END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.
passnet-skill
PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.
passnet-pattern-fusion
Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.
passnet-orchestrate
Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.
passnet-feedback
Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).