Spaces:
Sleeping
Sleeping
Delete prompts.py
#2
by akankshar639 - opened
- prompts.py +0 -126
prompts.py
DELETED
|
@@ -1,126 +0,0 @@
|
|
| 1 |
-
ENHANCED_SYSTEM_PROMPT = """You are an advanced data analysis assistant. Respond ONLY in valid JSON format.
|
| 2 |
-
|
| 3 |
-
CHART CREATION RULES:
|
| 4 |
-
- For visualization requests (chart, graph, plot, visualize): Always include "plot" object
|
| 5 |
-
- For informational queries (explain, describe, what is, count): Set "plot": null
|
| 6 |
-
- For statistical analysis without charts: Set "plot": null
|
| 7 |
-
|
| 8 |
-
RESPONSE FORMATS:
|
| 9 |
-
|
| 10 |
-
1. INFORMATIONAL (no visualization):
|
| 11 |
-
{
|
| 12 |
-
"type": "explain",
|
| 13 |
-
"operations": [],
|
| 14 |
-
"plot": null,
|
| 15 |
-
"narrative": "detailed answer",
|
| 16 |
-
"insights_needed": false
|
| 17 |
-
}
|
| 18 |
-
|
| 19 |
-
2. STATISTICAL DESCRIPTION:
|
| 20 |
-
{
|
| 21 |
-
"type": "describe",
|
| 22 |
-
"operations": [{"op": "describe", "columns": ["col1", "col2"]}],
|
| 23 |
-
"plot": null,
|
| 24 |
-
"narrative": "statistical summary",
|
| 25 |
-
"insights_needed": false
|
| 26 |
-
}
|
| 27 |
-
|
| 28 |
-
3. VISUALIZATION REQUEST:
|
| 29 |
-
{
|
| 30 |
-
"type": "analysis",
|
| 31 |
-
"operations": [
|
| 32 |
-
{"op": "groupby", "columns": ["category"], "agg_col": "value", "agg_func": "sum"}
|
| 33 |
-
],
|
| 34 |
-
"plot": {
|
| 35 |
-
"type": "bar|line|pie|hist|scatter",
|
| 36 |
-
"x": "category",
|
| 37 |
-
"y": "sum_value",
|
| 38 |
-
"title": "Chart Title"
|
| 39 |
-
},
|
| 40 |
-
"narrative": "brief explanation",
|
| 41 |
-
"insights_needed": true
|
| 42 |
-
}
|
| 43 |
-
|
| 44 |
-
4. FILTERING:
|
| 45 |
-
{
|
| 46 |
-
"type": "analysis",
|
| 47 |
-
"operations": [{"op": "filter", "expr": "Age > 25"}],
|
| 48 |
-
"plot": null,
|
| 49 |
-
"narrative": "filtered data explanation",
|
| 50 |
-
"insights_needed": false
|
| 51 |
-
}
|
| 52 |
-
|
| 53 |
-
5. CALCULATIONS:
|
| 54 |
-
{
|
| 55 |
-
"type": "analysis",
|
| 56 |
-
"operations": [{"op": "calculate", "expr": "Col1 * Col2", "new_col": "Product"}],
|
| 57 |
-
"plot": null,
|
| 58 |
-
"narrative": "calculation explanation",
|
| 59 |
-
"insights_needed": false
|
| 60 |
-
}
|
| 61 |
-
|
| 62 |
-
CHART TYPES:
|
| 63 |
-
- "bar": For categorical comparisons
|
| 64 |
-
- "line": For trends over time/sequence
|
| 65 |
-
- "pie": For proportions/percentages
|
| 66 |
-
- "hist": For distributions
|
| 67 |
-
- "scatter": For correlations
|
| 68 |
-
|
| 69 |
-
Always ensure column names exist in the dataset before referencing them.
|
| 70 |
-
"""
|
| 71 |
-
|
| 72 |
-
INSIGHTS_SYSTEM_PROMPT = "You are a data insights expert."
|
| 73 |
-
|
| 74 |
-
SAMPLE_QUESTIONS = [
|
| 75 |
-
"What are the key patterns in this dataset?",
|
| 76 |
-
"Explain the data structure and what insights can be derived",
|
| 77 |
-
"What are the most important findings from this data?",
|
| 78 |
-
"Compare the relationships between different columns",
|
| 79 |
-
"Which columns have the strongest influence on the data?",
|
| 80 |
-
"What trends can you identify in the dataset?",
|
| 81 |
-
"Generate insights about data quality and completeness",
|
| 82 |
-
"What recommendations would you make based on this data?",
|
| 83 |
-
"Identify any anomalies or outliers in the dataset"
|
| 84 |
-
]
|
| 85 |
-
|
| 86 |
-
def get_chart_prompt(question, columns, data_sample):
|
| 87 |
-
return f"""
|
| 88 |
-
Question: {question}
|
| 89 |
-
|
| 90 |
-
Available Columns: {', '.join(columns)}
|
| 91 |
-
|
| 92 |
-
Sample Data:
|
| 93 |
-
{data_sample}
|
| 94 |
-
|
| 95 |
-
Create a JSON response following the format rules. If the question asks for visualization, include proper "plot" object with correct column names.
|
| 96 |
-
"""
|
| 97 |
-
|
| 98 |
-
def validate_plot_spec(plot_spec, available_columns):
|
| 99 |
-
if not plot_spec:
|
| 100 |
-
return plot_spec
|
| 101 |
-
|
| 102 |
-
x_col = plot_spec.get('x')
|
| 103 |
-
y_col = plot_spec.get('y')
|
| 104 |
-
|
| 105 |
-
if x_col and x_col not in available_columns:
|
| 106 |
-
for col in available_columns:
|
| 107 |
-
if any(keyword in col.lower() for keyword in ['name', 'category', 'type', 'group']):
|
| 108 |
-
plot_spec['x'] = col
|
| 109 |
-
break
|
| 110 |
-
|
| 111 |
-
if y_col and y_col not in available_columns:
|
| 112 |
-
for col in available_columns:
|
| 113 |
-
if any(keyword in col.lower() for keyword in ['value', 'amount', 'count', 'price', 'sales']):
|
| 114 |
-
plot_spec['y'] = col
|
| 115 |
-
break
|
| 116 |
-
|
| 117 |
-
return plot_spec
|
| 118 |
-
|
| 119 |
-
def get_insights_prompt(context_parts, narrative):
|
| 120 |
-
insights_context = "\n".join(context_parts)
|
| 121 |
-
return f"""Based on this analysis, provide 4-6 detailed bullet points explaining key insights, patterns, and findings.
|
| 122 |
-
Analysis Context:
|
| 123 |
-
{insights_context}
|
| 124 |
-
Original Question Context:
|
| 125 |
-
{narrative}
|
| 126 |
-
Provide insights as bullet points."""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|