app.py

#run from app.py
import streamlit as st
st.set_page_config(layout="wide")
import streamlit.components.v1 as components
import numpy as np
import pandas as pd
import os
import sys
import plotly.figure_factory as ff
import plotly.express as px
import matplotlib.pyplot as plt
####################### UTILITY FUNCTIONS ####################
def convert_df(frame):#-------------------converts a dataframe to csv data
	return frame.to_csv().encode('utf-8')

def showresults(frame, file): #-----------displays data frame and lets you download it.
	#HtmlFile = open("data.html", 'r', encoding='utf-8')
	#source_code = HtmlFile.read()
	#components.html(source_code, height = 775,width=900)
	sf = frame.astype(str)
	st.dataframe(sf)
	csv = convert_df(frame)
	csvname = file + ".csv"
	keyname = file + 'download-csv'
	st.download_button(
	   "Press to Download",
	   csv,
	   csvname,
	   "text/csv",
	   key=keyname
	)
def mergeframes(leftframe, rightframe, column):#---merges two tables based on the  column
	#mergeframe = pd.merge(leftframe, rightframe, on=["Username"])
	mergeframe = pd.merge(rightframe, leftframe, how="outer", on=column)
#	st.write(mergeframe.shape)
	return(mergeframe)
def comboframes(leftframe, rightframe, column):#---merges two tables based on the column
	comboframe = pd.merge(rightframe, leftframe, how="outer", on=column)
#	st.write(mergeframe.shape)
	return(comboframe)

####################### MAIN ####################
def main():
	st.markdown('## Data App')
#---------------------------Sidebar--------------------------
	todo = st.sidebar.selectbox("Do you want to:",
		['<select>', "Enhance Confluence Data", "Merge Two Files", "Beta"])

#---------------------------Sidebar and Page Logic --------------------------
	if todo == "<select>":
		st.write("Use the sidebar to the left to select about views or edits of Confluence pages.")

	if todo == "Enhance Confluence Data":
		#seldtype = st.radio("Viewtracker data type:", ('View Data', 'Edit Data',))
		selectviz = st.sidebar.radio("Select a visualization", ('Stacked Bar Graph', 'Organizational Sunburst', 'Parallel Categories', 'Three Dimensional Scatterplot', 'Beta',))
		cfile = st.file_uploader("Upload a Confluence Viewtracker file")
		ofile = st.file_uploader("Upload another data file")
		fixUsername = st.checkbox('Update Username column?', value=True,)
		if cfile is not None and ofile is not None:
			#dataframe = pd.read_csv(uploaded_file)
			otherdata = pd.read_csv(ofile)
			confdata = pd.read_csv(cfile)
			if fixUsername:
				st.write('Updated Username column to show full email')
				confdata['Username'] = confdata['Username'].astype(str)  + "@salesforce.com"
			st.write("Enter the name of the column used to combine the files:")
			colname = st.text_input('Column name')
			# if not string or if string best way to test for string existence instead of if not None
			if colname:
				vf = mergeframes(otherdata, confdata, colname)

				#drop null values, because otherwise the data can't be used.
				UseIncomplete = st.sidebar.radio("Use incomplete data?", ('Yes','No')) #---------------Use Incomplete Data or Not?
				if UseIncomplete == 'No':
					vf = vf.dropna()
				else:
					vf = vf.fillna(value='Unknown')
					graphs(vf, selectviz, todo)
		else:
			st.write("Select a file")
#---------------------------Combine Files --------------------------
	if todo == "Merge Two Files":
		st.markdown('### Combine two files')
		uploaded_files = st.file_uploader("Step 1: select two files to combine.", accept_multiple_files=True)
		if len(uploaded_files) < 2 or len(uploaded_files) > 2:
			st.write("Upload exactly 2 files")
		if len(uploaded_files) == 2:
			file0 = pd.read_csv(uploaded_files[0])
			file1 = pd.read_csv(uploaded_files[1])

			st.write("Enter the name of the column used to combine the files:")
			combo = st.text_input('Column name')
			if combo:
#				st.write("Combine on column: " + combo)
				cf = comboframes(file0, file1, combo)
				#View the data?
				option = st.selectbox('Do you want to edit, view, or visualize data?',
				('View and Download Data', 'Edit Data','Visualize Data'))
			#-------------EDIT FILES
			if option == 'Edit Data':
				#Select a file to edit
				file0name = uploaded_files[0].name
				file1name = uploaded_files[1].name
				# Finding Common columns
				#a = np.intersect1d(df2.columns, df1.columns)

			# Printing common columns
			#st.write("Common Columns:",a)
				editfilename = st.selectbox('Select a file to edit',
				(uploaded_files[0].name, uploaded_files[1].name))
#				Replace a column name
				st.write(editfilename)
				st.write("Replace a column name:")
				if editfilename == file0name:
					st.write("Use " + file0name)
				if editfilename == file1name:
					st.write("Use " + file1name)
#				df.rename(columns = {'old_col1':'new_col1'}, inplace = True)
#				Append a string to a column
#				Remove lines from a file
			#-------------VIEW AND DOWNLOAD
			if option == 'View and Download Data':
				viewfile0 = st.checkbox("Uploaded Files", value=True,)
				viewcombofile = st.checkbox('Combined File', value=True,)
			#	st.write(mergeframe.shape)
				if viewfile0:
					st.write("Filename:", uploaded_files[0].name)
					st.write("Records and columns:", file0.shape)
					#st.write(file0)
					showresults(file0, uploaded_files[0].name)
					st.write("Filename:", uploaded_files[1].name)
					st.write("Records and columns:", file1.shape)
					#st.write(file1)
					showresults(file1, uploaded_files[1].name)
				if viewcombofile:
					st.write("Filename: combo.csv")
					st.write("Records and columns:", cf.shape)
					#st.dataframe(cf)
					showresults(cf, "combinedfile")
			#-------------VISUALIZE DATA
			if option == 'Visualize Data':
				comboviz = st.sidebar.radio("Select a visualization", ('Stacked Bar Graph', 'Organizational Sunburst', 'Parallel Categories'))
				graphs(cf, comboviz, "combofile")
#		for uploaded_file in uploaded_files:
#			#bytes_data = uploaded_file.read()
#			st.write("filename:", uploaded_file.name)
#			dataframe = pd.read_csv(uploaded_file)
#			st.write(dataframe)
#			# df.get("firstname")
#--------------------SELECT GRAPH TYPE-------------------------
def graphs(df, selectviz, todo):
	#Get unique values for user input

	#spaces = df['Space Name'].unique()
	#orgs = df['Cost Center'].unique()
	#roles = df['Role'].unique()
	if selectviz == "Stacked Bar Graph":
		st.header( selectviz + " for Views")
		viewstackedbartitle = "Views By "
		stackedbar(df, viewstackedbartitle)
		showresults(df, todo)
	elif selectviz == "Organizational Sunburst": #-------------SUNBURST
		st.write("Organizational Sunburst")
		sunburst(df)
		showresults(df, todo)
	elif selectviz == "Parallel Categories": #-----------------PARALLEL CATEGORIES
		st.write("Parallel Categories")
		parallelCategories(df)
	elif selectviz == "Beta": #--------------------------------Beta
		st.write("Combined Raw Data")
		st.dataframe(df)
		#assign a unique ID for all users based on username
		df['User ID'] = df.groupby(['Username']).ngroup()

		st.write("Anonymous Raw Data")
		st.dataframe(df)
		st.write("Download Anonymous View Data")
		#showresults(dfff, ref, physics)
		st.write("Anonymous Edit Data")
#		detail(ef)
## search type input
#	searchterm = st.sidebar.text_input('Enter a keyword', value="", max_chars=25)
#	search = st.sidebar.radio(
#		"Keyword search for:",
#		('Role', 'Manager', 'Page', 'Space'))


####################### RENDERING FUNCTIONS ####################
#---------------------------Stacked Bar--------------------------
def stackedbar(frame, title):
	#st.markdown("# Select view comparisons")
	collist = list(frame)
	colsort  = np.sort(collist)
	colbar = colsort
	colseg = colsort
	col1, col2= st.columns(2)
	with col1:
		st.markdown('**Bars:**')
#		st.write(collist)
		xselect = st.selectbox('Select data for bar graph:', colbar)
		cselect = st.selectbox('Select data for bar segments:', colseg)
	with col2:
		st.markdown('**Filters:**')
#		st.markdown('**Colors:**')
		fselect = st.selectbox('Select data to filter on:', colbar)
		st.write("Enter the name of the column used to combine the files:")
		filtertext = st.text_input('Enter text to filter on')
		#col = "account"
		#df1[col].isin(df2[col].values)
	writerfilter = st.radio("Remove writers and editors?:", ('View Data', 'Edit Data',))
	fig = px.histogram(frame, x=xselect, color=cselect)
	fig.update_layout(
		height=600,
		width=1000
	)
	st.plotly_chart(fig, use_container_width=True)

#---------------------------Sunburst--------------------------
def sunburst(frame):
	# insert at index 0 column Visits with value 1
	sdf=frame.assign(Views=1)
	#st.dataframe(sdf)

	fig1 = px.sunburst(sdf, path=['Country', 'State/Province', 'City'], values='Views',
		color='Country',
		color_continuous_scale="RdYlGn",
		width=750, height=750
	)
	fig2 = px.sunburst(sdf, path=['Mgmt Chain - Level 02', 'Mgmt Chain - Level 03', 'Mgmt Chain - Level 04', 'Mgmt Chain - Level 05'],
		values='Views',
		color='Mgmt Chain - Level 05',
		color_continuous_scale='RdBu',
		width=750, height=750
	)
	fig3 = px.sunburst(sdf, path=['Space Name', 'Content Type', 'Visit ID', 'Content Title'],
		values='Views',
		color='Space Name',
		color_continuous_scale='RdBu',
		width=750, height=750
	)
	st.markdown('## By Country, State, and City')
	st.plotly_chart(fig1, use_container_width=True)
	st.markdown('## By Management Chain')
	st.plotly_chart(fig2, use_container_width=True)
#	st.markdown('**Views by Space Name, Content Type, Visit ID, and Content Title**')
#	st.plotly_chart(fig3, use_container_width=True)
#---------------------------Parallel Categories--------------------------
def parallelCategories(frame):
	st.dataframe(frame)
	collist = list(frame)
	colsort  = np.sort(collist)
	colbar = colsort
	colseg = colsort
	col1, col2= st.columns(2)
	with col1:
#		st.markdown('**Bars:**')
#		st.write(collist)
		xselect = st.multiselect('Select Categories:', colbar)
		fframe = frame.loc[:, xselect]
	with col2:
		clrlist = list(fframe)
		clrsort  = np.sort(clrlist)
		clrbar = clrsort

#		st.markdown('**Colors:**')
		cselect = st.selectbox('Select Color:', clrbar)

		#Add a count of unique values for the color scale
		#frame['Counts'] = frame.groupby(['Title'])['Space Key'].transform('count')
		fframe['Counts'] = fframe.groupby(by=[cselect]).transform('count')

	#fig = px.parallel_categories(frame)

	fig = px.parallel_categories(fframe, color="Counts", color_continuous_scale=px.colors.sequential.Rainbow)
	st.plotly_chart(fig, use_container_width=True)

if __name__ == "__main__":
	main()