data-visualization

Name: data-visualization
Author: langchain-ai

27kstars3.7kforks

How to install data-visualization?

langchain-ai/deepagents/data-visualization

$npx -y skills add langchain-ai/deepagents --skill data-visualization

Installs into the current project.

Summary

Use for creating publication-quality charts and multi-panel analysis summaries. Triggers when tasks involve visualizing data, plotting results, creating charts, or producing visual reports from analysis output.

Files · 1

View on GitHub

SKILL.md

1	# Data Visualization Skill
2
3	Create publication-quality analytical charts using matplotlib and seaborn in a headless GPU sandbox. Charts are saved as PNG files to `/workspace/` for retrieval.
4
5	## When to Use This Skill
6
7	Use this skill when:
8	- Visualizing results from cuDF analysis or cuML models
9	- Creating charts (bar, line, scatter, heatmap, histogram, box plot)
10	- Building multi-panel analysis summaries
11	- The user asks for visual output, plots, graphs, or charts
12	- Presenting statistical findings with figures
13
14	## Initialization (REQUIRED)
15
16	MUST call `matplotlib.use('Agg')` BEFORE importing pyplot. This enables headless rendering.
17
18	```python
19	import matplotlib
20	matplotlib.use('Agg') # Headless backend — MUST be before pyplot import
21	import matplotlib.pyplot as plt
22	import numpy as np
23
24	# Publication-quality defaults
25	plt.rcParams.update({
26	'figure.dpi': 100,
27	'savefig.dpi': 300,
28	'font.size': 11,
29	'axes.labelsize': 12,
30	'axes.titlesize': 14,
31	'xtick.labelsize': 10,
32	'ytick.labelsize': 10,
33	'legend.fontsize': 10,
34	'figure.constrained_layout.use': True,
35	})
36
37	# Colorblind-safe palette (Okabe-Ito)
38	COLORS = ['#0173B2', '#DE8F05', '#029E73', '#D55E00', '#CC78BC',
39	'#CA9161', '#FBAFE4', '#949494', '#ECE133', '#56B4E9']
40	```
41
42	## Saving Charts
43
44	Always save to `/workspace/` with these settings:
45
46	```python
47	plt.savefig('/workspace/chart_name.png', dpi=300, bbox_inches='tight',
48	facecolor='white', edgecolor='none')
49	plt.close()
50	# IMPORTANT: call read_file("/workspace/<chart>.png") to display inline
51	```
52
53	- `dpi=300` for print quality
54	- `bbox_inches='tight'` removes excess whitespace
55	- `facecolor='white'` ensures white background
56	- Always call `plt.close()` after saving to free memory
57
58	## Displaying Charts (REQUIRED)
59
60	After saving any chart, you MUST call `read_file` on it to display it inline in the conversation:
61
62	```
63	read_file("/workspace/chart_name.png")
64	```
65
66	Users cannot see charts unless you do this. Every chart you save MUST be followed by a `read_file` call.
67
68	## Quick Reference
69
70	### Bar Chart (from groupby results)
71
72	```python
73	# After: result = to_pd(df.groupby("category")["value"].mean())
74	fig, ax = plt.subplots(figsize=(8, 5))
75
76	bars = ax.bar(result.index, result.values, color=COLORS[:len(result)],
77	edgecolor='black', linewidth=0.8)
78
79	for bar in bars:
80	height = bar.get_height()
81	ax.text(bar.get_x() + bar.get_width()/2., height,
82	f'{height:.1f}', ha='center', va='bottom', fontsize=9)
83
84	ax.set_ylabel('Mean Value', fontweight='bold')
85	ax.set_xlabel('Category', fontweight='bold')
86	ax.set_title('Average Value by Category', fontweight='bold')
87	ax.grid(axis='y', alpha=0.3, linestyle='--')
88	ax.set_axisbelow(True)
89
90	plt.savefig('/workspace/bar_chart.png', dpi=300, bbox_inches='tight',
91	facecolor='white', edgecolor='none')
92	plt.close()
93	# IMPORTANT: call read_file("/workspace/<chart>.png") to display inline
94	```
95
96	### Line Chart (trends over time)
97
98	```python
99	fig, ax = plt.subplots(figsize=(10, 5))
100
101	for i, col in enumerate(columns_to_plot):
102	ax.plot(df["date"], df[col], label=col, color=COLORS[i], linewidth=2,
103	marker='o', markersize=3, markevery=max(1, len(df)//20))
104
105	ax.set_ylabel('Values', fontweight='bold')
106	ax.set_xlabel('Date', fontweight='bold')
107	ax.set_title('Trends Over Time', fontweight='bold')
108	ax.legend(frameon=True, shadow=False)
109	ax.grid(True, alpha=0.3, linestyle='--')
110	ax.set_axisbelow(True)
111	plt.xticks(rotation=45, ha='right')
112
113	plt.savefig('/workspace/line_chart.png', dpi=300, bbox_inches='tight',
114	facecolor='white', edgecolor='none')
115	plt.close()
116	# IMPORTANT: call read_file("/workspace/<chart>.png") to display inline
117	```
118
119	### Scatter Plot — Continuous Color (correlations)
120
121	```python
122	fig, ax = plt.subplots(figsize=(8, 6))
123
124	scatter = ax.scatter(df["x"], df["y"], c=df["value"], cmap='viridis',
125	s=40, alpha=0.7, edgecolors='black', linewidth=0.3)
126	plt.colorbar(scatter, ax=ax, label='Value')
127
128	# Optional: trend line
129	z = np.polyfit(df["x"], df["y"], 1)
130	ax.plot(df["x"].sort_values(), np.poly1d(z)(df["x"].sort_values()),
131	"r--", linewidth=2, label=f'y={z[0]:.2f}x+{z[1]:.2f}')
132
133	ax.set_xlabel('X', fontweight='bold')
134	ax.set_ylabel('Y', fontweight='bold')
135	ax.set_title('Correlation Analysis', fontweight='bold')
136	ax.legend()
137	ax.grid(True, alpha=0.3, linestyle='--')
138
139	plt.savefig('/workspace/scatter_correlation.png', dpi=300, bbox_inches='tight',
140	facecolor='white', edgecolor='none')
141	plt.close()
142	# IMPORTANT: call read_file("/workspace/<chart>.png") to display inline
143	```
144
145	### Scatter Plot — Categorical Color (clusters)
146
147	```python
148	fig, ax = plt.subplots(figsize=(8, 6))
149
150	for i, label in enumerate(sorted(df["cluster"].unique())):
151	mask = df["cluster"] == l

People who liked this also viewed

6 picks

Type

data-visualization

27kstars3.7kforks

View on GitHub

How to install data-visualization?

langchain-ai/deepagents/data-visualization

$npx -y skills add langchain-ai/deepagents --skill data-visualization

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Data Visualization Skill
2
3	Create publication-quality analytical charts using matplotlib and seaborn in a headless GPU sandbox. Charts are saved as PNG files to `/workspace/` for retrieval.
4
5	## When to Use This Skill
6
7	Use this skill when:
8	- Visualizing results from cuDF analysis or cuML models
9	- Creating charts (bar, line, scatter, heatmap, histogram, box plot)
10	- Building multi-panel analysis summaries
11	- The user asks for visual output, plots, graphs, or charts
12	- Presenting statistical findings with figures
13
14	## Initialization (REQUIRED)
15
16	MUST call `matplotlib.use('Agg')` BEFORE importing pyplot. This enables headless rendering.
17
18	```python
19	import matplotlib
20	matplotlib.use('Agg') # Headless backend — MUST be before pyplot import
21	import matplotlib.pyplot as plt
22	import numpy as np
23
24	# Publication-quality defaults
25	plt.rcParams.update({
26	'figure.dpi': 100,
27	'savefig.dpi': 300,
28	'font.size': 11,
29	'axes.labelsize': 12,
30	'axes.titlesize': 14,
31	'xtick.labelsize': 10,
32	'ytick.labelsize': 10,
33	'legend.fontsize': 10,
34	'figure.constrained_layout.use': True,
35	})
36
37	# Colorblind-safe palette (Okabe-Ito)
38	COLORS = ['#0173B2', '#DE8F05', '#029E73', '#D55E00', '#CC78BC',
39	'#CA9161', '#FBAFE4', '#949494', '#ECE133', '#56B4E9']
40	```
41
42	## Saving Charts
43
44	Always save to `/workspace/` with these settings:
45
46	```python
47	plt.savefig('/workspace/chart_name.png', dpi=300, bbox_inches='tight',
48	facecolor='white', edgecolor='none')
49	plt.close()
50	# IMPORTANT: call read_file("/workspace/<chart>.png") to display inline
51	```
52
53	- `dpi=300` for print quality
54	- `bbox_inches='tight'` removes excess whitespace
55	- `facecolor='white'` ensures white background
56	- Always call `plt.close()` after saving to free memory
57
58	## Displaying Charts (REQUIRED)
59
60	After saving any chart, you MUST call `read_file` on it to display it inline in the conversation:
61
62	```
63	read_file("/workspace/chart_name.png")
64	```
65
66	Users cannot see charts unless you do this. Every chart you save MUST be followed by a `read_file` call.
67
68	## Quick Reference
69
70	### Bar Chart (from groupby results)
71
72	```python
73	# After: result = to_pd(df.groupby("category")["value"].mean())
74	fig, ax = plt.subplots(figsize=(8, 5))
75
76	bars = ax.bar(result.index, result.values, color=COLORS[:len(result)],
77	edgecolor='black', linewidth=0.8)
78
79	for bar in bars:
80	height = bar.get_height()
81	ax.text(bar.get_x() + bar.get_width()/2., height,
82	f'{height:.1f}', ha='center', va='bottom', fontsize=9)
83
84	ax.set_ylabel('Mean Value', fontweight='bold')
85	ax.set_xlabel('Category', fontweight='bold')
86	ax.set_title('Average Value by Category', fontweight='bold')
87	ax.grid(axis='y', alpha=0.3, linestyle='--')
88	ax.set_axisbelow(True)
89
90	plt.savefig('/workspace/bar_chart.png', dpi=300, bbox_inches='tight',
91	facecolor='white', edgecolor='none')
92	plt.close()
93	# IMPORTANT: call read_file("/workspace/<chart>.png") to display inline
94	```
95
96	### Line Chart (trends over time)
97
98	```python
99	fig, ax = plt.subplots(figsize=(10, 5))
100
101	for i, col in enumerate(columns_to_plot):
102	ax.plot(df["date"], df[col], label=col, color=COLORS[i], linewidth=2,
103	marker='o', markersize=3, markevery=max(1, len(df)//20))
104
105	ax.set_ylabel('Values', fontweight='bold')
106	ax.set_xlabel('Date', fontweight='bold')
107	ax.set_title('Trends Over Time', fontweight='bold')
108	ax.legend(frameon=True, shadow=False)
109	ax.grid(True, alpha=0.3, linestyle='--')
110	ax.set_axisbelow(True)
111	plt.xticks(rotation=45, ha='right')
112
113	plt.savefig('/workspace/line_chart.png', dpi=300, bbox_inches='tight',
114	facecolor='white', edgecolor='none')
115	plt.close()
116	# IMPORTANT: call read_file("/workspace/<chart>.png") to display inline
117	```
118
119	### Scatter Plot — Continuous Color (correlations)
120
121	```python
122	fig, ax = plt.subplots(figsize=(8, 6))
123
124	scatter = ax.scatter(df["x"], df["y"], c=df["value"], cmap='viridis',
125	s=40, alpha=0.7, edgecolors='black', linewidth=0.3)
126	plt.colorbar(scatter, ax=ax, label='Value')
127
128	# Optional: trend line
129	z = np.polyfit(df["x"], df["y"], 1)
130	ax.plot(df["x"].sort_values(), np.poly1d(z)(df["x"].sort_values()),
131	"r--", linewidth=2, label=f'y={z[0]:.2f}x+{z[1]:.2f}')
132
133	ax.set_xlabel('X', fontweight='bold')
134	ax.set_ylabel('Y', fontweight='bold')
135	ax.set_title('Correlation Analysis', fontweight='bold')
136	ax.legend()
137	ax.grid(True, alpha=0.3, linestyle='--')
138
139	plt.savefig('/workspace/scatter_correlation.png', dpi=300, bbox_inches='tight',
140	facecolor='white', edgecolor='none')
141	plt.close()
142	# IMPORTANT: call read_file("/workspace/<chart>.png") to display inline
143	```
144
145	### Scatter Plot — Categorical Color (clusters)
146
147	```python
148	fig, ax = plt.subplots(figsize=(8, 6))
149
150	for i, label in enumerate(sorted(df["cluster"].unique())):
151	mask = df["cluster"] == l

People who liked this also viewed

6 picks

Type