documentations/lte_drop_traffic_doc.py

import streamlit as st

st.markdown(
    """
# LTE Drop Traffic Documentation

## 1. Objective
Detect LTE cells with significant traffic drop by comparing recent days against long-term average.

## 2. When to use this tool
Use this page for early detection of degraded LTE traffic behavior and location-based prioritization.

## 3. Input files and accepted formats
- Required: traffic report in `.csv`
- Current parser expects `;` delimiter.

## 4. Required columns/fields
Input CSV should contain at least:
- `PERIOD_START_TIME` (format `%m.%d.%Y`)
- `LNCEL name`
- `4G/LTE DL Traffic Volume (GBytes)`
- `4G/LTE UL Traffic Volume (GBytes)`

## 5. Step-by-step usage
1. Open `KPI Analysis > LTE Drop Traffic Analysis`.
2. Upload CSV report.
3. Set number of last days and drop threshold.
4. Review affected cells table.
5. Explore trend plot by selected cell.
6. Review map and download affected cells.

## 6. Outputs generated
- affected cells dataframe with `drop_%`
- downloadable file `traffic_drop_cells.xlsx`
- trend plot per selected cell
- map of dropped cells

## 7. Frequent errors and fixes
- Date parsing error.
  - Fix: ensure `PERIOD_START_TIME` matches `%m.%d.%Y` format.
- Empty affected cells list.
  - Fix: lower threshold or increase analysis window.
- Missing map coordinates.
  - Fix: ensure physical DB mapping by code is available.

## 8. Minimal reproducible example
- Input: LTE traffic CSV with several days and measurable drop on some cells.
- Action: set threshold to `50%` and run analysis.
- Expected result: non-empty affected cell list and downloadable Excel.

## 9. Known limitations
- Uses mean-based long-term vs recent-day comparison.
- Relies on code matching against physical database for map.
- Strict date format requirement in input parser.

## 10. Version and update date
- Documentation version: 1.0
- Last update: 2026-02-23
"""
)