app.py

import streamlit as st
import pandas as pd
import plotly.express as px
from streamlit_option_menu import option_menu
from numerize.numerize import numerize
import time
from streamlit_extras.metric_cards import style_metric_cards
st.set_option('deprecation.showPyplotGlobalUse', False)
import plotly.graph_objs as go

#uncomment this line if you use mysql
#from query import *
#Creating session variables
    
st.set_page_config(page_title="Dashboard",page_icon="🌍",layout="wide")
st.header("Data Analytics, KPI, Trends & Predictions")

df = pd.DataFrame()

uploaded_file = st.file_uploader(
    "Choose a CSV file. This should be in long format (one datapoint per row).",
    type="csv",
    )
if uploaded_file is not None:
    df = pd.read_csv(uploaded_file)

else:
    #load excel file | comment this line when  you fetch data from mysql
    df=pd.read_excel('data.xlsx', sheet_name='Sheet1')# load Style css

with open('style.css')as f:
    st.markdown(f"<style>{f.read()}</style>", unsafe_allow_html = True)
    
if "openai_key" not in st.session_state:
    with st.form("API key"):
        key = st.text_input("OpenAI Key", value="", type="password")
        if st.form_submit_button("Submit"):
            st.session_state.openai_key = key
            st.success('Saved API key for this session.')


#all graphs we use custom css not streamlit 
theme_plotly = None 

#uncomment these two lines if you fetch data from mysql
#result = view_all_data()
#df=pd.DataFrame(result,columns=["StaffNum","Engage","Location","State","Region","Staff","Job","BusinessType","Absentee","Training","Rating","id"])


#side bar logo

#switcher

region=st.sidebar.multiselect(
    "SELECT REGION",
     options=df["Region"].unique(),
     default=df["Region"].unique(),
)
location=st.sidebar.multiselect(
    "SELECT LOCATION",
     options=df["Location"].unique(),
     default=df["Location"].unique(),
)
construction=st.sidebar.multiselect(
    "SELECT BUSINESS TYPE",
     options=df["BusinessType"].unique(),
     default=df["BusinessType"].unique(),
)
df_selection=df.query(
    "Region==@region & Location==@location"
)

#this function performs basic descriptive analytics like Mean,Mode,Sum  etc
def Home():
    with st.expander("VIEW EXCEL DATASET"):
        showData=st.multiselect('Filter: ',df_selection.columns,default=["StaffNum","Engage","Location","State","Region","Salary","Job","BusinessType","Absentee","Training","Rating"])
        st.dataframe(df_selection[showData],use_container_width=True)

    #compute top analytics
    total_salary = float(pd.Series(df_selection['Salary']).sum())
    salary_mode = float(pd.Series(df_selection['Salary']).mode())
    salary_mean = float(pd.Series(df_selection['Salary']).mean())
    salary_median= float(pd.Series(df_selection['Salary']).median()) 
    rating = float(pd.Series(df_selection['Rating']).sum())


    total1,total2,total3,total4,total5=st.columns(5,gap='small')
    with total1:
        st.info('Sum Salary',icon="💰")
        st.metric(label="Sum TZS",value=f"{total_salary:,.0f}")

    with total2:
        st.info('Most Salary',icon="💰")
        st.metric(label="Mode TZS",value=f"{salary_mode:,.0f}")

    with total3:
        st.info('Average Salary',icon="💰")
        st.metric(label="Average TZS",value=f"{salary_mean:,.0f}")

    with total4:
        st.info('Central Earnings',icon="💰")
        st.metric(label="Median TZS",value=f"{salary_median:,.0f}")

    with total5:
        st.info('Ratings',icon="💰")
        st.metric(label="Rating",value=numerize(rating),help=f""" Total Rating: {rating} """)
    style_metric_cards(background_color="#FFFFFF",border_left_color="#686664",border_color="#000000",box_shadow="#F71938")

    #variable distribution Histogram
    with st.expander("DISTRIBUTIONS BY FREQUENCY"):
     df.hist(figsize=(16,8),color='#898784', zorder=2, rwidth=0.9,legend = ['Salary']);
     st.pyplot()

#graphs
def graphs():
    #total_salary=int(df_selection["Salary"]).sum()
    #averageRating=int(round(df_selection["Rating"]).mean(),2) 
    #simple bar graph salary by business type
    salary_by_business_type=(
        df_selection.groupby(by=["BusinessType"]).count()[["Salary"]].sort_values(by="Salary")
    )
    fig_investment=px.bar(
       salary_by_business_type,
       x="Salary",
       y=salary_by_business_type.index,
       orientation="h",
       title="<b> Total Salary BY BUSINESS TYPE </b>",
       color_discrete_sequence=["#0083B8"]*len(salary_by_business_type),
       template="plotly_white",
    )
    fig_investment.update_layout(
     plot_bgcolor="rgba(0,0,0,0)",
     font=dict(color="black"),
     yaxis=dict(showgrid=True, gridcolor='#cecdcd'),  # Show y-axis grid and set its color  
     paper_bgcolor='rgba(0, 0, 0, 0)',  # Set paper background color to transparent
     xaxis=dict(showgrid=True, gridcolor='#cecdcd'),  # Show x-axis grid and set its color
     )

    #simple line graph investment by state
    investment_state=df_selection.groupby(by=["State"]).count()[["Salary"]]
    fig_state=px.line(
       investment_state,
       x=investment_state.index,
       y="Salary",
       orientation="v",
       title="<b> Total Salary BY STATE </b>",
       color_discrete_sequence=["#0083b8"]*len(investment_state),
       template="plotly_white",
    )
    fig_state.update_layout(
    xaxis=dict(tickmode="linear"),
    plot_bgcolor="rgba(0,0,0,0)",
    yaxis=(dict(showgrid=False))
     )

    left,right,center=st.columns(3)
    left.plotly_chart(fig_state,use_container_width=True)
    right.plotly_chart(fig_investment,use_container_width=True)
    
    with center:
      #pie chart
      fig = px.pie(df_selection, values='Rating', names='State', title='RATINGS BY REGIONS')
      fig.update_layout(legend_title="Regions", legend_y=0.9)
      fig.update_traces(textinfo='percent+label', textposition='inside')
      st.plotly_chart(fig, use_container_width=True, theme=theme_plotly)

#function to show current earnings against expected target     
def Progressbar():
    st.markdown("""<style>.stProgress > div > div > div > div { background-image: linear-gradient(to right, #99ff99 , #FFFF00)}</style>""",unsafe_allow_html=True,)
    target=3000000000
    current=df_selection["Salary"].sum()
    percent=round((current/target*100))
    mybar=st.progress(0)

    if percent>100:
        st.subheader("Target done !")
    else:
     st.write("you have ",percent, "% " ,"of ", (format(target, 'd')), "TZS")
     for percent_complete in range(percent):
        time.sleep(0.1)
        mybar.progress(percent_complete+1,text=" Target Percentage")

#menu bar
def sideBar():
 with st.sidebar:
    selected=option_menu(
        menu_title="Main Menu",
        options=["Home","Progress"],
        icons=["house","eye"],
        menu_icon="cast",
        default_index=0
    )
 if selected=="Home":
    #st.subheader(f"Page: {selected}")
    Home()
    graphs()
 if selected=="Progress":
    #st.subheader(f"Page: {selected}")
    Progressbar()
    graphs()

sideBar()
st.sidebar.image("data/logo1.png",caption="")


st.subheader('Pick Features to Explore Distribution Tredns by Quartiles',)
#feature_x = st.selectbox('Select feature for x Qualitative data', df_selection.select_dtypes("object").columns)
feature_y = st.selectbox('Select feature for y Quantitative Data', df_selection.select_dtypes("number").columns)
fig2 = go.Figure(
    data=[go.Box(x=df['BusinessType'], y=df[feature_y])],
    layout=go.Layout(
        title=go.layout.Title(text="Business Type by Quartiles Salary"),
        plot_bgcolor='rgba(0, 0, 0, 0)',  # Set plot background color to transparent
        paper_bgcolor='rgba(0, 0, 0, 0)',  # Set paper background color to transparent
        xaxis=dict(showgrid=True, gridcolor='#cecdcd'),  # Show x-axis grid and set its color
        yaxis=dict(showgrid=True, gridcolor='#cecdcd'),  # Show y-axis grid and set its color
        font=dict(color='#cecdcd'),  # Set text color to black
    )
)
# Display the Plotly figure using Streamlit
st.plotly_chart(fig2,use_container_width=True)



#theme
hide_st_style=""" 

<style>
#MainMenu {visibility:hidden;}
footer {visibility:hidden;}
header {visibility:hidden;}
</style>
"""