File size: 8,446 Bytes
0898fcc |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 |
from typing import List, Tuple
import pandas as pd
import plotly.express as px
import streamlit as st
from queries.process_gsm import combined_gsm_database
from utils.convert_to_excel import convert_gsm_dfs, save_dataframe
from utils.kpi_analysis_utils import create_hourly_date, kpi_naming_cleaning
# Constants
GSM_COLUMNS = [
"ID_BTS",
"BSC",
"code",
"Region",
"locationAreaIdLAC",
"Longitude",
"Latitude",
]
TRX_COLUMNS = [
"ID_BTS",
"number_trx_per_cell",
"number_tch_per_cell",
"number_sd_per_cell",
"number_bcch_per_cell",
"number_ccch_per_cell",
"number_cbc_per_cell",
"number_total_channels_per_cell",
"number_signals_per_cell",
]
KPI_COLUMNS = [
"BSC_name",
"BCF_name",
"BTS_name",
"Paging_messages_on_air_interface",
"DELETE_PAGING_COMMAND_c003038",
"datetime",
"date",
"hour",
]
def get_gsm_databases(dump_path: str) -> pd.DataFrame:
"""
Process GSM database dump and return combined DataFrame with BTS and TRX data.
Args:
dump_path: Path to the GSM dump file
Returns:
pd.DataFrame: Combined DataFrame with BTS and TRX information
"""
dfs = combined_gsm_database(dump_path)
bts_df: pd.DataFrame = dfs[0]
trx_df: pd.DataFrame = dfs[2]
# Clean GSM df
bts_df = bts_df[GSM_COLUMNS]
trx_df = trx_df[TRX_COLUMNS]
trx_df = trx_df.drop_duplicates(subset=["ID_BTS"])
gsm_df = pd.merge(bts_df, trx_df, on="ID_BTS", how="left")
# Create BSC_Lac column
gsm_df["BSC_Lac"] = (
gsm_df["BSC"].astype(str) + "_" + gsm_df["locationAreaIdLAC"].astype(str)
)
# Calculate number of TRX per LAC
gsm_df["number_trx_per_lac"] = gsm_df.groupby("BSC_Lac")[
"number_trx_per_cell"
].transform("sum")
return gsm_df
def analyze_lac_load_kpi(hourly_report_path: str) -> pd.DataFrame:
"""
Process hourly KPI report and prepare it for LAC load analysis.
Args:
hourly_report_path: Path to the hourly KPI report CSV file
Returns:
pd.DataFrame: Processed DataFrame with KPI data
"""
df = pd.read_csv(hourly_report_path, delimiter=";")
df = kpi_naming_cleaning(df)
df = create_hourly_date(df)
df = df[KPI_COLUMNS]
# Clean and process BTS codes
df = df[df["BTS_name"].str.len() >= 5]
df["code"] = df["BTS_name"].str.split("_").str[0]
df["code"] = pd.to_numeric(df["code"], errors="coerce").fillna(0).astype(int)
return df
def analyze_lac_load(dump_path: str, hourly_report_path: str) -> List[pd.DataFrame]:
"""
Analyze LAC load from GSM dump and hourly KPI report.
Args:
dump_path: Path to the GSM dump file
hourly_report_path: Path to the hourly KPI report CSV file
Returns:
List containing two DataFrames: [lac_load_df, max_paging_df]
"""
gsm_df = get_gsm_databases(dump_path)
lac_load_df = analyze_lac_load_kpi(hourly_report_path)
lac_load_df = pd.merge(gsm_df, lac_load_df, on="code", how="left")
# Aggregate data
lac_load_df = (
lac_load_df.groupby(
[
"datetime",
"date",
"hour",
"BSC_name",
"BSC_Lac",
"number_trx_per_lac",
]
)
.agg(
{
"Paging_messages_on_air_interface": "max",
"DELETE_PAGING_COMMAND_c003038": "max",
}
)
.reset_index()
)
# Get max paging messages
max_paging_messages = lac_load_df.sort_values(
by=["BSC_Lac", "Paging_messages_on_air_interface"], ascending=False
).drop_duplicates(subset=["BSC_Lac"], keep="first")[
[
"BSC_name",
"BSC_Lac",
"number_trx_per_lac",
"Paging_messages_on_air_interface",
]
]
# Get max delete paging commands
max_delete_paging = lac_load_df.sort_values(
by=["BSC_Lac", "DELETE_PAGING_COMMAND_c003038"], ascending=False
).drop_duplicates(subset=["BSC_Lac"], keep="first")[
["BSC_name", "BSC_Lac", "DELETE_PAGING_COMMAND_c003038"]
]
# Merge results
max_paging_df = pd.merge(
max_paging_messages,
max_delete_paging,
on=["BSC_name", "BSC_Lac"],
how="left",
)
# Calculate utilization (paging/640800)
max_paging_df["Utilization"] = (
(max_paging_df["Paging_messages_on_air_interface"] / 640800) * 100
).round(2)
return [lac_load_df, max_paging_df]
def display_ui() -> None:
"""Display the Streamlit user interface."""
st.title(" 📊 GSM LAC Load Analysis")
doc_col, image_col = st.columns(2)
with doc_col:
st.write(
"""
The report should be run with a minimum of 7 days of data.
- Dump file required
- Hourly Report in CSV format
"""
)
with image_col:
st.image("./assets/gsm_lac_load.png", width=250)
@st.fragment
def display_filtered_lac_load(lac_load_df: pd.DataFrame) -> None:
"""
Display filtered LAC load data with interactive charts.
Args:
lac_load_df: DataFrame containing LAC load data
"""
st.write("### Filtered LAC Load by BSC and BSC_Lac")
bsc_col, bsc_lac_col = st.columns(2)
with bsc_col:
selected_bsc = st.multiselect(
"Select BSC",
lac_load_df["BSC_name"].unique(),
key="selected_bsc",
default=[lac_load_df["BSC_name"].unique()[0]],
)
with bsc_lac_col:
selected_bsc_lac = st.multiselect(
"Select BSC_Lac",
lac_load_df[lac_load_df["BSC_name"].isin(selected_bsc)]["BSC_Lac"].unique(),
key="selected_bsc_lac",
default=lac_load_df[lac_load_df["BSC_name"].isin(selected_bsc)][
"BSC_Lac"
].unique(),
)
filtered_lac_load_df = lac_load_df[
lac_load_df["BSC_name"].isin(selected_bsc)
& lac_load_df["BSC_Lac"].isin(selected_bsc_lac)
]
# Display charts
chart1, chart2 = st.columns(2)
with chart1:
st.write("### Paging Messages on Air Interface")
fig1 = px.line(
filtered_lac_load_df,
x="datetime",
y="Paging_messages_on_air_interface",
color="BSC_Lac",
title="Max Paging Messages on Air Interface Per BSC_Lac",
)
fig1.update_layout(
xaxis_title="<b>Datetime</b>",
yaxis_title="<b>Paging Messages on Air Interface</b>",
)
fig1.add_hline(y=256000, line_color="red", line_dash="dash", line_width=2)
st.plotly_chart(fig1)
with chart2:
st.write("### Delete Paging Commands")
fig2 = px.line(
filtered_lac_load_df,
x="datetime",
y="DELETE_PAGING_COMMAND_c003038",
color="BSC_Lac",
title="Max Delete Paging Commands Per BSC_Lac",
)
fig2.update_layout(
xaxis_title="<b>Datetime</b>",
yaxis_title="<b>Delete Paging Commands</b>",
)
st.plotly_chart(fig2)
st.write("### Filtered LAC Load Data")
st.dataframe(filtered_lac_load_df)
def main() -> None:
"""Main function to run the Streamlit app."""
display_ui()
# File uploaders
file1, file2 = st.columns(2)
with file1:
uploaded_dump = st.file_uploader("Upload Dump file in xlsb format", type="xlsb")
with file2:
uploaded_hourly_report = st.file_uploader(
"Upload Hourly Report in CSV format", type="csv"
)
if uploaded_dump is not None and uploaded_hourly_report is not None:
if st.button("Analyze Data", type="primary"):
with st.spinner("Analyzing data..."):
dfs = analyze_lac_load(
dump_path=uploaded_dump,
hourly_report_path=uploaded_hourly_report,
)
lac_load_df = dfs[0]
max_paging_df = dfs[1]
if lac_load_df is not None and "lac_load_df" not in st.session_state:
st.session_state.lac_load_df = lac_load_df
st.write("### LAC Load and Utilization with Max Paging 640800")
st.dataframe(max_paging_df)
display_filtered_lac_load(lac_load_df)
if __name__ == "__main__":
main()
|