statistical-analysis

Name: statistical-analysis
Author: anthropics

3.3kinstalls2.7kforks

How to install statistical-analysis?

anthropics/knowledge-work-plugins/statistical-analysis

$npx -y skills add anthropics/knowledge-work-plugins --skill statistical-analysis

Installs into the current project.

Summary

statistical-analysis is an agent skill published from anthropics/knowledge-work-plugins, installable through the skills CLI.

Files · 1

View on GitHub

SKILL.md

1	# Statistical Analysis Skill
2
3	Descriptive statistics, trend analysis, outlier detection, hypothesis testing, and guidance on when to be cautious about statistical claims.
4
5	## Descriptive Statistics Methodology
6
7	### Central Tendency
8
9	Choose the right measure of center based on the data:
10
11	\| Situation \| Use \| Why \|
12	\|---\|---\|---\|
13	\| Symmetric distribution, no outliers \| Mean \| Most efficient estimator \|
14	\| Skewed distribution \| Median \| Robust to outliers \|
15	\| Categorical or ordinal data \| Mode \| Only option for non-numeric \|
16	\| Highly skewed with outliers (e.g., revenue per user) \| Median + mean \| Report both; the gap shows skew \|
17
18	Always report mean and median together for business metrics. If they diverge significantly, the data is skewed and the mean alone is misleading.
19
20	### Spread and Variability
21
22	- Standard deviation: How far values typically fall from the mean. Use with normally distributed data.
23	- Interquartile range (IQR): Distance from p25 to p75. Robust to outliers. Use with skewed data.
24	- Coefficient of variation (CV): StdDev / Mean. Use to compare variability across metrics with different scales.
25	- Range: Max minus min. Sensitive to outliers but gives a quick sense of data extent.
26
27	### Percentiles for Business Context
28
29	Report key percentiles to tell a richer story than mean alone:
30
31	```
32	p1: Bottom 1% (floor / minimum typical value)
33	p5: Low end of normal range
34	p25: First quartile
35	p50: Median (typical user)
36	p75: Third quartile
37	p90: Top 10% / power users
38	p95: High end of normal range
39	p99: Top 1% / extreme users
40	```
41
42	Example narrative: "The median session duration is 4.2 minutes, but the top 10% of users spend over 22 minutes per session, pulling the mean up to 7.8 minutes."
43
44	### Describing Distributions
45
46	Characterize every numeric distribution you analyze:
47
48	- Shape: Normal, right-skewed, left-skewed, bimodal, uniform, heavy-tailed
49	- Center: Mean and median (and the gap between them)
50	- Spread: Standard deviation or IQR
51	- Outliers: How many and how extreme
52	- Bounds: Is there a natural floor (zero) or ceiling (100%)?
53
54	## Trend Analysis and Forecasting
55
56	### Identifying Trends
57
58	Moving averages to smooth noise:
59	```python
60	# 7-day moving average (good for daily data with weekly seasonality)
61	df['ma_7d'] = df['metric'].rolling(window=7, min_periods=1).mean()
62
63	# 28-day moving average (smooths weekly AND monthly patterns)
64	df['ma_28d'] = df['metric'].rolling(window=28, min_periods=1).mean()
65	```
66
67	Period-over-period comparison:
68	- Week-over-week (WoW): Compare to same day last week
69	- Month-over-month (MoM): Compare to same month prior
70	- Year-over-year (YoY): Gold standard for seasonal businesses
71	- Same-day-last-year: Compare specific calendar day
72
73	Growth rates:
74	```
75	Simple growth: (current - previous) / previous
76	CAGR: (ending / beginning) ^ (1 / years) - 1
77	Log growth: ln(current / previous) -- better for volatile series
78	```
79
80	### Seasonality Detection
81
82	Check for periodic patterns:
83	1. Plot the raw time series -- visual inspection first
84	2. Compute day-of-week averages: is there a clear weekly pattern?
85	3. Compute month-of-year averages: is there an annual cycle?
86	4. When comparing periods, always use YoY or same-period comparisons to avoid conflating trend with seasonality
87
88	### Forecasting (Simple Methods)
89
90	For business analysts (not data scientists), use straightforward methods:
91
92	- Naive forecast: Tomorrow = today. Use as a baseline.
93	- Seasonal naive: Tomorrow = same day last week/year.
94	- Linear trend: Fit a line to historical data. Only for clearly linear trends.
95	- Moving average forecast: Use trailing average as the forecast.
96
97	Always communicate uncertainty. Provide a range, not a point estimate:
98	- "We expect 10K-12K signups next month based on the 3-month trend"
99	- NOT "We will get exactly 11,234 signups next month"
100
101	When to escalate to a data scientist: Non-linear trends, multiple seasonalities, external factors (marketing spend, holidays), or when forecast accuracy matters for resource allocation.
102
103	## Outlier and Anomaly Detection
104
105	### Statistical Methods
106
107	Z-score method (for normally distributed data):
108	```python
109	z_scores = (df['value'] - df['value'].mean()) / df['value'].std()
110	outliers = df[abs(z_scores) > 3] # More than 3 standard deviations
111	```
112
113	IQR method (robust to non-normal distributions):
114	```python
115	Q1 = df['value'].quantile(0.25)
116	Q3 = df['value'].quantile(0.75)
117	IQR = Q3 - Q1
118	lower_bound = Q1 - 1.5 * IQR
119	upper_bound = Q3 + 1.5 * IQR
120	outliers = df[(df['value'] < lower_bound) \| (df['value'] > upper_bound)]
121	```
122
123	Percentile method (simplest):
124	```python
125	outliers = df[(df['value'] < df['value'].qua

People who liked this also viewed

6 picks

Type

lark-drive飞书云空间（云盘/云存储）：管理 Drive 文件和文件夹，包含上传/下载、创建文件夹、复制/移动/删除、查看元数据、评论/权限/订阅、标题、版本、飞书文档密级标签（secure labels）和本地文件导入。用户需要整理云盘目录、处理云空间资源 URL/token、判断链接类型/真实 token/标题，或导入 Word/Markdown/Excel/CSV/PPTX/.base 为 docx/sSkillsJul 2026382k16k
lark-sharedUse for lark-cli setup/auth tasks: auth login/status/logout, user vs bot identity, business-domain permissions (--domain, including all/docs/drive), missing scopes, revoking authorization, or handlingSkillsJul 2026382k16k
lark-whiteboard飞书画板：查询和编辑飞书云文档中的画板。支持导出画板为预览图片、导出原始节点结构、使用多种格式更新画板内容。当用户需要查看画板内容、导出画板图片、编辑画板时使用此 skill。不负责：飞书云文档内容编辑（lark-doc）、文档内嵌电子表格/Base（lark-sheets / lark-base）。SkillsJul 2026380k16k
lark-task飞书任务：管理任务、清单和任务智能体。创建待办任务、查看和更新任务状态、拆分子任务、组织任务清单、分配协作成员、上传任务附件、注册或注销任务智能体、更新任务智能体的主页数据、写入智能体任务记录。当用户需要创建待办事项、查看任务列表、跟踪任务进度、管理项目清单或给他人分配任务、为任务上传附件文件、注册注销任务智能体、更新智能体主页数据、写入任务记录时使用。SkillsJul 2026380k16k
lark-attendance飞书考勤打卡：查询自己的考勤打卡记录SkillsJul 2026330k16k
lark-apps妙搭（Spark/Miaoda）应用开发与托管：应用创建、本地全栈开发、云端生成迭代、创意设计（UI mockup / 可交互原型 / 线框图 / 落地页 / 仪表盘 / 幻灯片 deck / 视觉探索）、AI相关能力和飞书平台能力或者其他外部能力集成、日志/Trace/监控指标/PV/UV 查询、环境变量管理、应用角色与成员管理、自动化触发器（定时/记录变更/Webhook/飞书审批）。当用户要SkillsJul 2026227k16k

statistical-analysis

3.3kinstalls2.7kforks

View on GitHub

How to install statistical-analysis?

anthropics/knowledge-work-plugins/statistical-analysis

$npx -y skills add anthropics/knowledge-work-plugins --skill statistical-analysis

Installs into the current project.

Summary

statistical-analysis is an agent skill published from anthropics/knowledge-work-plugins, installable through the skills CLI.

Files · 1

View on GitHub

SKILL.md

1	# Statistical Analysis Skill
2
3	Descriptive statistics, trend analysis, outlier detection, hypothesis testing, and guidance on when to be cautious about statistical claims.
4
5	## Descriptive Statistics Methodology
6
7	### Central Tendency
8
9	Choose the right measure of center based on the data:
10
11	\| Situation \| Use \| Why \|
12	\|---\|---\|---\|
13	\| Symmetric distribution, no outliers \| Mean \| Most efficient estimator \|
14	\| Skewed distribution \| Median \| Robust to outliers \|
15	\| Categorical or ordinal data \| Mode \| Only option for non-numeric \|
16	\| Highly skewed with outliers (e.g., revenue per user) \| Median + mean \| Report both; the gap shows skew \|
17
18	Always report mean and median together for business metrics. If they diverge significantly, the data is skewed and the mean alone is misleading.
19
20	### Spread and Variability
21
22	- Standard deviation: How far values typically fall from the mean. Use with normally distributed data.
23	- Interquartile range (IQR): Distance from p25 to p75. Robust to outliers. Use with skewed data.
24	- Coefficient of variation (CV): StdDev / Mean. Use to compare variability across metrics with different scales.
25	- Range: Max minus min. Sensitive to outliers but gives a quick sense of data extent.
26
27	### Percentiles for Business Context
28
29	Report key percentiles to tell a richer story than mean alone:
30
31	```
32	p1: Bottom 1% (floor / minimum typical value)
33	p5: Low end of normal range
34	p25: First quartile
35	p50: Median (typical user)
36	p75: Third quartile
37	p90: Top 10% / power users
38	p95: High end of normal range
39	p99: Top 1% / extreme users
40	```
41
42	Example narrative: "The median session duration is 4.2 minutes, but the top 10% of users spend over 22 minutes per session, pulling the mean up to 7.8 minutes."
43
44	### Describing Distributions
45
46	Characterize every numeric distribution you analyze:
47
48	- Shape: Normal, right-skewed, left-skewed, bimodal, uniform, heavy-tailed
49	- Center: Mean and median (and the gap between them)
50	- Spread: Standard deviation or IQR
51	- Outliers: How many and how extreme
52	- Bounds: Is there a natural floor (zero) or ceiling (100%)?
53
54	## Trend Analysis and Forecasting
55
56	### Identifying Trends
57
58	Moving averages to smooth noise:
59	```python
60	# 7-day moving average (good for daily data with weekly seasonality)
61	df['ma_7d'] = df['metric'].rolling(window=7, min_periods=1).mean()
62
63	# 28-day moving average (smooths weekly AND monthly patterns)
64	df['ma_28d'] = df['metric'].rolling(window=28, min_periods=1).mean()
65	```
66
67	Period-over-period comparison:
68	- Week-over-week (WoW): Compare to same day last week
69	- Month-over-month (MoM): Compare to same month prior
70	- Year-over-year (YoY): Gold standard for seasonal businesses
71	- Same-day-last-year: Compare specific calendar day
72
73	Growth rates:
74	```
75	Simple growth: (current - previous) / previous
76	CAGR: (ending / beginning) ^ (1 / years) - 1
77	Log growth: ln(current / previous) -- better for volatile series
78	```
79
80	### Seasonality Detection
81
82	Check for periodic patterns:
83	1. Plot the raw time series -- visual inspection first
84	2. Compute day-of-week averages: is there a clear weekly pattern?
85	3. Compute month-of-year averages: is there an annual cycle?
86	4. When comparing periods, always use YoY or same-period comparisons to avoid conflating trend with seasonality
87
88	### Forecasting (Simple Methods)
89
90	For business analysts (not data scientists), use straightforward methods:
91
92	- Naive forecast: Tomorrow = today. Use as a baseline.
93	- Seasonal naive: Tomorrow = same day last week/year.
94	- Linear trend: Fit a line to historical data. Only for clearly linear trends.
95	- Moving average forecast: Use trailing average as the forecast.
96
97	Always communicate uncertainty. Provide a range, not a point estimate:
98	- "We expect 10K-12K signups next month based on the 3-month trend"
99	- NOT "We will get exactly 11,234 signups next month"
100
101	When to escalate to a data scientist: Non-linear trends, multiple seasonalities, external factors (marketing spend, holidays), or when forecast accuracy matters for resource allocation.
102
103	## Outlier and Anomaly Detection
104
105	### Statistical Methods
106
107	Z-score method (for normally distributed data):
108	```python
109	z_scores = (df['value'] - df['value'].mean()) / df['value'].std()
110	outliers = df[abs(z_scores) > 3] # More than 3 standard deviations
111	```
112
113	IQR method (robust to non-normal distributions):
114	```python
115	Q1 = df['value'].quantile(0.25)
116	Q3 = df['value'].quantile(0.75)
117	IQR = Q3 - Q1
118	lower_bound = Q1 - 1.5 * IQR
119	upper_bound = Q3 + 1.5 * IQR
120	outliers = df[(df['value'] < lower_bound) \| (df['value'] > upper_bound)]
121	```
122
123	Percentile method (simplest):
124	```python
125	outliers = df[(df['value'] < df['value'].qua

People who liked this also viewed

6 picks

Type