prismWriter 1.5.0

Python library for programmatically creating and manipulating GraphPad Prism (.pzfx) files

prismWriter

Python library and GUI for programmatically creating and manipulating GraphPad Prism (.pzfx) files. Convert pandas DataFrames into Prism's XML-based table format with complex grouping structures.

Version: 1.0.0

Features

• Convert pandas DataFrames to GraphPad Prism format
• Flexible grouping: main groups, sub-columns, row labels
• Multiple interfaces: Python API, Qt GUI, and Streamlit Web interface
• Load and modify existing .pzfx files
• Validate table structure with preview functionality
• Automatic backup creation when loading existing files

Installation

pip install -e .

Requirements

• Python >= 3.7
• numpy
• pandas
• openpyxl (for Excel file support)
• PySide6 (for Qt GUI) [optional]
• streamlit >= 1.28.0 (for web interface) [optional]

See Also:

• https://github.com/Yue-Jiang/pypzfx - Another python package with similar functionality
• https://github.com/Biomiha/prism2R - R package for reading .prism (newer) files
• https://github.com/yue-jiang/pzfx - R package for reading .pzfx files

Quick Start

Python API

from prismWriter.prism_writer import PrismFile
import pandas as pd

# Create a new PrismFile
pf = PrismFile()

# Load your data
df = pd.read_csv('data.csv')

# Create a grouped table
pf.make_group_table(
    group_name="MyTable",
    group_values=df,
    groupby="condition",  # Main grouping column
    cols=["value1", "value2"],  # Data columns
    subgroupby="replicate"  # Sub-grouping
)

# Save to Prism file
pf.save("output.pzfx")

Web Interface (Streamlit)

The easiest way to use prismWriter:

streamlit run prismWriter/streamlit_app.py

Or use the provided launch script:

• Windows: run_web.bat

Or after installation:

prismwriter-web

Features:

• Upload CSV/Excel files (with multi-sheet Excel support)
• Live data preview with automatic column type detection
• Interactive grouping configuration
• Preview table structure before generating
• Direct download of generated .pzfx files
• Load and view existing Prism files

Qt GUI

python prismWriter/gui.py

Or after installation:

prismwriter-gui

API Reference

PrismFile Class

pf = PrismFile(file=None)  # Create new or load existing file

Methods:

• load(file_path, backup=True) - Load an existing Prism file
• smake_group_table(group_name, group_values, groupby=None, cols=None, subgroupby=None, rowgroupby=None, append=True) - Create a grouped table
• get_table_names() - Get list of all table names
• to_dataframe(table_name) - Convert a table to pandas DataFrame
• save(file_path) - Save to file
• write(file_path, xml_declaration=True, encoding='utf-8', pretty_print=True) - Write with options
• remove_table(table_name) - Remove a table from the file

Grouping Parameters

• groupby: Main column grouping - creates separate Y-columns for each category
• subgroupby: Sub-columns within each Y-column (string for column name, or list for data columns)
• rowgroupby: Row labels on the vertical axis (string for column name, or list for data columns)
• cols: Data columns to include

Note: Use either column name (subgroupby='condition') OR list of data columns (subgroupby=['col1', 'col2']) - not both.

Example Data Structure

import pandas as pd
from prismWriter.prism_writer import PrismFile

# Sample data
data = {
    'treatment': ['Control', 'Control', 'Drug A', 'Drug A'],
    'time': [0, 24, 0, 24],
    'replicate': [1, 1, 1, 1],
    'value': [10.5, 15.2, 12.1, 25.8]
}
df = pd.DataFrame(data)

# Create Prism file
pf = PrismFile()
pf.make_group_table(
    group_name="Treatment_Data",
    group_values=df,
    groupby="treatment",    # Separate columns for Control vs Drug A
    rowgroupby="time",      # Rows labeled by timepoint
    cols=["value"]          # Data column
)
pf.save("treatment_data.pzfx")

Loading Existing Files

from prismWriter.prism_writer import PrismFile

# Load existing Prism file (creates automatic backup)
pf = PrismFile()
pf.load("existing.pzfx")

# Or load directly in constructor
pf = PrismFile(file="existing.pzfx")

# Get table names
tables = pf.get_table_names()
print(f"Found tables: {tables}")

# Convert table to DataFrame
df = pf.to_dataframe("TableName")

# Add new table
pf.make_group_table(
    group_name="NewTable",
    group_values=df,
    groupby="category",
    cols=["value"]
)

# Save (use write() for more options)
pf.save("modified.pzfx")

Dependencies

• Core: pandas, numpy
• GUI: PySide6 (Qt bindings)
• Web: streamlit >= 1.28.0
• Files: openpyxl for Excel support

Project Structure

prismWriter/
├── prismWriter/
│   ├── __init__.py          # Package exports
│   ├── prism_writer.py      # Core XML generation engine
│   ├── gui.py               # PySide6 Qt GUI wrapper
│   ├── streamlit_app.py     # Streamlit web interface
│   ├── prism_template2.pzfx # XML structure template
│   └── schema/              # Prism XML schemas
├── tests/
│   ├── test_prism_writer.py
│   ├── test_gui.py
│   └── test_data.csv
├── pyproject.toml
└── README.md

Development

Running Tests

pytest tests/

Key Implementation Details

• Uses prism_template2.pzfx as XML structure blueprint
• XML Namespace: {http://graphpad.com/prism/Prism.htm}
• "Raveling" strategy: transforms DataFrames into flat structure via melt(), then rebuilds into nested XML hierarchy
• Automatic backup files created as {filename}.backup{timestamp} when loading existing files
• Logging enabled at INFO level for debugging table creation

Known Limitations

• No validation for duplicate table names
• Row mismatch warnings can occur with sparse data
• Template system hardcoded to specific Prism XML schema version
• No undo functionality - backup files are the safety net

Roadmap

• Add support for other table types (X-Y, nested, Multi var?)
• Add support for .prism loading (maybe saving?)

License

MIT License

Contributing

Contributions welcome! Please feel free to submit a Pull Request.