# imports

from bokeh.models import ColumnDataSource

from bokeh.plotting import figure, output_file, show

from bokeh.models.widgets import Panel, Tabs

from bokeh.layouts import column, row, layout, widgetbox

from bokeh.models import CustomJS, Slider, Toggle, Dropdown, MultiChoice, Spinner

from bokeh.transform import dodge

from bokeh.palettes import RdBu as colors

from bokeh.models import ColorBar, LinearColorMapper, HoverTool, BoxSelectTool

import pandas as pd

import bisect

from math import pi

from numpy import arange

from itertools import chain

from collections import OrderedDict

# why is 'plot' created??? I can't find any place that add any data to it... (see p1)

plot = figure(tools="pan,wheel_zoom,box_zoom,reset") # create figure object

plot.add_tools(BoxSelectTool(dimensions="width")) # add interactive tools to plot

output_file("test.html") # set output path

df = pd.read_excel('archive\dataset.xlsx') # dataset path (relative to script)

#tab 1 - Assignment 4 visualisation

# !LOOK! the age_quantiles are binned per 4 entries... BUT ITS HARDCODED!!!

# what if we want to increase or decreate the binsize? than we need to add or remove a lot of if/elif statements

# we can put it in a loop to simplify it (see below)

# consider using the following aproach:

'''

binnumber = 5 # number of bins

binsize = int(20/binnumber) # number of age_quantiles in one bin

# lists to store number of people per category per bin in

regularWardPos = [0]*binsize # creates a list of zeroes of length 'binsize'

semiIntensivePos = [0]*binsize

intensivePos = [0]*binsize

regularWardNeg = [0]*binsize

semiIntensiveNeg = [0]*binsize

intensiveNeg = [0]*binsize

totalAgePos = [0]*binsize

totalAgeNeg = [0]*binsize

'''

regularWardPos = [0]*5 # list of zeroes for each bin and type of patient (see second visualisation for explanetion of list of zeroes)

semiIntensivePos = [0]*5

intensivePos = [0]*5

regularWardNeg = [0]*5

semiIntensiveNeg = [0]*5

intensiveNeg = [0]*5

totalAgePos = [0]*5

totalAgeNeg = [0]*5

for index, row in df.iterrows(): # iterate over dataframe rows

# extracting number of patients per age group per ward per test result

# The bins follow a pattern, which can thus be put into a loop to prevent repetition, replace 93 lines of code with 26:

'''

for bin in binnumber: # iterate over the bins

binStartAge = bin * binsize # computes at which age quantile a bin starts

# checks if age_quantile at 'index' is in the current bin:

if binStartAge <= df.iloc[index, 1] and df.iloc[index, 1] < (binStartAge + binnumber):

if df.iloc[index, 2] == "positive": # if patient is positive...

totalAgePos[bin] += 1 # ... increment counter for number of people in current bin that are positive

# identify at which ward the *positive* patient is in and increment corresponding counter

if df.iloc[index, 3] == 1:

regularWardPos[bin] += 1

elif df.iloc[index, 4] == 1:

semiIntensivePos[bin] += 1

elif df.iloc[index, 5] == 1:

intensivePos[bin] += 1

if df.iloc[index, 2] == "negative": # if patient is negative...

totalAgeNeg[bin] += 1 # ... increment counter for number of people in current bin that are negative

# identify at which ward the *negative* patient is in and increment corresponding counter

if df.iloc[index, 3] == 1:

regularWardNeg[bin] += 1

elif df.iloc[index, 4] == 1:

semiIntensiveNeg[bin] += 1

elif df.iloc[index, 5] == 1:

intensiveNeg[bin] += 1

'''

if df.iloc[index, 1] == 0 or df.iloc[index, 1] == 1 or df.iloc[index, 1] == 2 or df.iloc[index, 1] == 3:

if df.iloc[index, 2] == "positive":

totalAgePos[0] += 1

if df.iloc[index, 3] == 1:

regularWardPos[0] += 1

elif df.iloc[index, 4] == 1:

semiIntensivePos[0] += 1

elif df.iloc[index, 5] == 1:

intensivePos[0] += 1

if df.iloc[index, 2] == "negative":

totalAgeNeg[0] += 1

if df.iloc[index, 3] == 1:

regularWardNeg[0] += 1

elif df.iloc[index, 4] == 1:

semiIntensiveNeg[0] += 1

elif df.iloc[index, 5] == 1:

intensiveNeg[0] += 1

elif df.iloc[index, 1] == 4 or df.iloc[index, 1] == 5 or df.iloc[index, 1] == 6 or df.iloc[index, 1] == 7:

if df.iloc[index, 2] == "positive":

totalAgePos[1] += 1

if df.iloc[index, 3] == 1:

regularWardPos[1] += 1

elif df.iloc[index, 4] == 1:

semiIntensivePos[1] += 1

elif df.iloc[index, 5] == 1:

intensivePos[1] += 1

if df.iloc[index, 2] == "negative":

totalAgeNeg[1] += 1

if df.iloc[index, 3] == 1:

regularWardNeg[1] += 1

elif df.iloc[index, 4] == 1:

semiIntensiveNeg[1] += 1

elif df.iloc[index, 5] == 1:

intensiveNeg[1] += 1

elif df.iloc[index, 1] == 8 or df.iloc[index, 1] == 9 or df.iloc[index, 1] == 10 or df.iloc[index, 1] == 11:

if df.iloc[index, 2] == "positive":

totalAgePos[2] += 1

if df.iloc[index, 3] == 1:

regularWardPos[2] += 1

elif df.iloc[index, 4] == 1:

semiIntensivePos[2] += 1

elif df.iloc[index, 5] == 1:

intensivePos[2] += 1

if df.iloc[index, 2] == "negative":

totalAgeNeg[2] += 1

if df.iloc[index, 3] == 1:

regularWardNeg[2] += 1

elif df.iloc[index, 4] == 1:

semiIntensiveNeg[2] += 1

elif df.iloc[index, 5] == 1:

intensiveNeg[2] += 1

elif df.iloc[index, 1] == 12 or df.iloc[index, 1] == 13 or df.iloc[index, 1] == 14 or df.iloc[index, 1] == 15:

if df.iloc[index, 2] == "positive":

totalAgePos[3] += 1

if df.iloc[index, 3] == 1:

regularWardPos[3] += 1

elif df.iloc[index, 4] == 1:

semiIntensivePos[3] += 1

elif df.iloc[index, 5] == 1:

intensivePos[3] += 1

if df.iloc[index, 2] == "negative":

totalAgeNeg[3] += 1

if df.iloc[index, 3] == 1:

regularWardNeg[3] += 1

elif df.iloc[index, 4] == 1:

semiIntensiveNeg[3] += 1

elif df.iloc[index, 5] == 1:

intensiveNeg[3] += 1

elif df.iloc[index, 1] == 16 or df.iloc[index, 1] == 17 or df.iloc[index, 1] == 18 or df.iloc[index, 1] == 19:

if df.iloc[index, 2] == "positive":

totalAgePos[4] += 1

if df.iloc[index, 3] == 1:

regularWardPos[4] += 1

elif df.iloc[index, 4] == 1:

semiIntensivePos[4] += 1

elif df.iloc[index, 5] == 1:

intensivePos[4] += 1

if df.iloc[index, 2] == "negative":

totalAgeNeg[4] += 1

if df.iloc[index, 3] == 1:

regularWardNeg[4] += 1

elif df.iloc[index, 4] == 1:

semiIntensiveNeg[4] += 1

elif df.iloc[index, 5] == 1:

intensiveNeg[4] += 1

# lists of zeroes to store percentages per bin per type

percentageRegularWardPos = [0]*5 # !LOOK! replaced [0,0,0,0,0] with [0]*5... the 5 shouldn't be hardcoded (change to binnumber??? (see above))

percentageSemiIntensivePos = [0]*5

percentageIntensivePos = [0]*5

percentageRegularWardNeg = [0]*5

percentageSemiIntensiveNeg = [0]*5

percentageIntensiveNeg = [0]*5

for i in range(len(regularWardPos)):

# compute percentages of positive and negative patients per age group per ward

percentageRegularWardPos[i] = regularWardPos[i] / totalAgePos[i] * 100

percentageSemiIntensivePos[i] = semiIntensivePos[i] / totalAgePos[i] * 100

percentageIntensivePos[i] = intensivePos[i] / totalAgePos[i] * 100

percentageRegularWardNeg[i] = regularWardNeg[i] / totalAgeNeg[i] * 100

percentageSemiIntensiveNeg[i] = semiIntensiveNeg[i] / totalAgeNeg[i] * 100

percentageIntensiveNeg[i] = intensiveNeg[i] / totalAgeNeg[i] * 100

# Legend and axis labels

wardDevision = ["regular ward", "semi-intensive unit", "intensive care"]

ageDevision = ["child and teen", "young adult and adult", "middle aged", "senior", "elderly"]

positiveReg = ["regular ward positive", "regular ward negative"]

positiveSemi = ["semi-intensive unit positive", "semi-intensive unit negative"]

positiveIntens = ["intensive care positive", "intensive care negative"]

# packaging data into dictionary format

dictDataReg = {

}

dictDataSemi = {

}

dictDataIntens = {

}

# converting dictionaries into datasources

sourceReg = ColumnDataSource(data=dictDataReg)

sourceSemi = ColumnDataSource(data=dictDataSemi)

sourceIntens = ColumnDataSource(data=dictDataIntens)

# colors for the bars

colorsReg = ["#2874A6", "#85C1E9"] # regular ward color [positive test, negative test]

colorsSemi = ["#B03A2E", "#F5B7B1"] # semi-intense ward color [positive test, negative test]

colorsIntens = ["#B7950B", "#F9E79F"] # intensive ward color [positive test, negative test]

p1 = figure(

)

#dodge('age group', -0.25, range=chart.x_range)

# adds stacked bars to the plot

p1.vbar_stack(

positiveReg, # sets y-axis for bars (using keys of source)

x=dodge('age group', -0.25, range=p1.x_range), # sets x-axis

width=0.2, # width of bars

source=dictDataReg, # datasource for the bars

color= colorsReg, # colors of the bars

legend_label=positiveReg # set legend label

)

p1.vbar_stack(

positiveSemi, # sets y-axis for bars (using keys of source)

x=dodge('age group', 0.0, range=p1.x_range), # sets x-axis

width=0.2, # width of bars

source=dictDataSemi, # datasource for the bars

color= colorsSemi, # colors of the bars

legend_label=positiveSemi # set legend label

)

p1.vbar_stack(

positiveIntens, # sets y-axis for bars (using keys of source)

x=dodge('age group', 0.25, range=p1.x_range), # sets x-axis

width=0.2, # width of bars

source=dictDataIntens, # datasource for the bars

color= colorsIntens, # colors of the bars

legend_label=positiveIntens # set legend label

)

# plot properties

p1.y_range.start = 0 # y-range starting point

p1.x_range.range_padding = 0.1 # additional space between x-axis values

p1.xgrid.grid_line_color = None # removes gridlines

p1.legend.location = "top_left" # sets location of Legend

p1.legend.orientation = "vertical" # sets orientation of the Legend (vertical -> entries below each other)

#Visualisation 2 - Bar chart - Assignment 3

# !LOOK! instead of putting 20 zeroes in a list manually, consider using a list comprehension or a lsit operation

'''

positiveAge = [0 for iter in range(0,20)] # list comprehension (puts 0 in the list as it loops 20 times)

positiveAge = [0]*20 # list operation (multiply [0] by 20 => [0]+[0]+...+[0] = [0, 0, ..., 0])

# make lists of zeroes to store number of patients in:

positiveAge = [0]*20

totalAge = [0]*20

percentageAge = [0]*20

'''

positiveAge = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

totalAge = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

percentageAge = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

for index, row in df.iterrows(): # iterate over dataset

# !LOOK!

# contains 'if/elif' statement for each age quantile... this is quite bad practise:

# say we want to change "positive" to "negative", then we need to change it 20 times!

# rule of thumb: if you're repeating an if statement with only one number that's changing, you can make it shorter!

# in this case: consider using the index of the 'totalAge' as a comparison, which would replace 80 lines of code with 4:

'''

for age_quantile in range(len(totalAge)): # loops over integers 0 to 19

totalAge[age_quantile] += 1 # increments counter for number of people of age 'age_quantile'

if df.iloc[index, 2] == "positive": # if patient is positive...

positiveAge[age_quantile] += 1 # ... increment counter for number of people of age 'age_quantile' that is positive

'''

if df.iloc[index, 1] == 0:

totalAge[0] += 1

if df.iloc[index, 2] == "positive":

positiveAge[0] += 1

elif df.iloc[index, 1] == 1:

totalAge[1] += 1

if df.iloc[index, 2] == "positive":

positiveAge[1] += 1

elif df.iloc[index, 1] == 2:

totalAge[2] += 1

if df.iloc[index, 2] == "positive":

positiveAge[2] += 1

elif df.iloc[index, 1] == 3:

totalAge[3] += 1

if df.iloc[index, 2] == "positive":

positiveAge[3] += 1

elif df.iloc[index, 1] == 4:

totalAge[4] += 1

if df.iloc[index, 2] == "positive":

positiveAge[4] += 1

elif df.iloc[index, 1] == 5:

totalAge[5] += 1

if df.iloc[index, 2] == "positive":

positiveAge[5] += 1

elif df.iloc[index, 1] == 6:

totalAge[6] += 1

if df.iloc[index, 2] == "positive":

positiveAge[6] += 1

elif df.iloc[index, 1] == 7:

totalAge[7] += 1

if df.iloc[index, 2] == "positive":

positiveAge[7] += 1

elif df.iloc[index, 1] == 8:

totalAge[8] += 1

if df.iloc[index, 2] == "positive":

positiveAge[8] += 1

elif df.iloc[index, 1] == 9:

totalAge[9] += 1

if df.iloc[index, 2] == "positive":

positiveAge[9] += 1

elif df.iloc[index, 1] == 10:

totalAge[10] += 1

if df.iloc[index, 2] == "positive":

positiveAge[10] += 1

elif df.iloc[index, 1] == 11:

totalAge[11] += 1

if df.iloc[index, 2] == "positive":

positiveAge[11] += 1

elif df.iloc[index, 1] == 12:

totalAge[12] += 1

if df.iloc[index, 2] == "positive":

positiveAge[12] += 1

elif df.iloc[index, 1] == 13:

totalAge[13] += 1

if df.iloc[index, 2] == "positive":

positiveAge[13] += 1

elif df.iloc[index, 1] == 14:

totalAge[14] += 1

if df.iloc[index, 2] == "positive":

positiveAge[14] += 1

elif df.iloc[index, 1] == 15:

totalAge[15] += 1

if df.iloc[index, 2] == "positive":

positiveAge[15] += 1

elif df.iloc[index, 1] == 16:

totalAge[16] += 1

if df.iloc[index, 2] == "positive":

positiveAge[16] += 1

elif df.iloc[index, 1] == 17:

totalAge[17] += 1

if df.iloc[index, 2] == "positive":

positiveAge[17] += 1

elif df.iloc[index, 1] == 18:

totalAge[18] += 1

if df.iloc[index, 2] == "positive":

positiveAge[18] += 1

elif df.iloc[index, 1] == 19:

totalAge[19] += 1

if df.iloc[index, 2] == "positive":

positiveAge[19] += 1

for age in range(len(positiveAge)): # compute percentage of positvely tested patients per age quantile

percentageAge[age] = positiveAge[age] / totalAge[age] *100

print(positiveAge) # debugging

print(totalAge) # debugging

print(percentageAge) # debugging

# !LOOK! list of labels for the x-axis of the plot

# theres a pattern in the numbers, so it can be simplified! Consider using a list comprehension:

# ageQuantile = [str(number) for number in range(20)] # generates a list of numbers from 0 to 19 in string format

ageQuantile = ["0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15", "16", "17", "18", "19"]

# create plot 2 object

p2 = figure(

)

# adds bars to the plot 2 object

p2.vbar(

x=ageQuantile, # specifies at which x-axis each bar should be placed

top=percentageAge, # specifies y-axis (height of the bars)

width=0.5 # specifies width of the bars

)

# plot 2 properties

p2.xgrid.grid_line_color = None # disables gridlines

p2.y_range.start = 0 # sets starting point for y-axis

#Visualisation 3 - Heat map - Assignment 3

dfVirus = df.iloc[:, 21:38] # copy virus related column from dataset

dfVirus["SARS-Cov-2 exam result"] = df["SARS-Cov-2 exam result"] # add covid-test results to virus copy of dataset

del dfVirus['Mycoplasma pneumoniae'] # removing virusses with close to none entries

for i in range(16): # iterate over dataframe

# replace human langue values for virusses by easy to work with 1 and 0 values

dfVirus.iloc[:, i] = dfVirus.iloc[:, i].replace(["not_detected"], 0)

dfVirus.iloc[:, i] = dfVirus.iloc[:, i].replace(["detected"], 1)

# replace human langue values for covid test result by easy to work with 1 and 0 values

dfVirus.iloc[:, 16] = dfVirus.iloc[:, 16].replace(["negative"], 0)

dfVirus.iloc[:, 16] = dfVirus.iloc[:, 16].replace(["positive"], 1)

correlation = dfVirus.corr() # create correlation object

colors = list(reversed(colors[11])) # we want an odd number to ensure 0 correlation is a distinct color

labels = dfVirus.columns # use the dataframe column names (Virusses) as labels for the plot

nlabels = len(labels) # specify number of labels

def get_bounds(n):

return top, bottom, left, right

def get_colors(corr_array, colors):

return color

# create heatmap plot object

p3 = figure(

)

# plot 3 properties

p3.xgrid.grid_line_color = None # disables x-gridlines

p3.ygrid.grid_line_color = None # disables y-gridlines

p3.xaxis.major_label_orientation = pi/4 # rotates x-labels by 45 degrees (pi/4 rad => 45 deg)

p3.yaxis.major_label_orientation = pi/4 # rotates y-labels by 45 degrees

top, bottom, left, right = get_bounds(nlabels) # creates sqaures for plot

color_list = get_colors(correlation.values.flatten(), colors) # makes list of colors for values

# specifies how a square in the heatmap looks

p3.quad(

top = top,

bottom = bottom,

left = left,

right = right,

line_color = 'white',

color = color_list

)

# Set ticks with labels

ticks = [tick+0.5 for tick in list(range(nlabels))]

tick_dict = OrderedDict([[tick, labels[ii]] for ii, tick in enumerate(ticks)])

# Create the correct number of ticks for each axis

p3.xaxis.ticker = ticks

p3.yaxis.ticker = ticks

# Override the labels

p3.xaxis.major_label_overrides = tick_dict

p3.yaxis.major_label_overrides = tick_dict

# Setup color bar

mapper = LinearColorMapper(palette=colors, low=-1, high=1)

color_bar = ColorBar(color_mapper=mapper, location=(0, 0))

p3.add_layout(color_bar, 'right')

#Spinner GUI

spinner = Spinner(title="Size", low=0, high=4, step=0.1, value=1, width=80)

#spinner.js_link('value', points.glyph, 'radius')

#Dropdown menu GUI

menu = [("Item 1", "item_1"), ("Item 2", "item_2"), None, ("Item 3", "item_3")]

dropdown = Dropdown(label="Dropdown button", button_type="warning", menu=menu)

dropdown.js_on_event("menu_item_click", CustomJS(code="console.log('dropdown: ' + this.item, this.toString())"))

#Toggle button GUI

toggle = Toggle(label="Button", button_type="success")

toggle.js_on_click(CustomJS(code="""

console.log('toggle: active=' + this.active, this.toString())

"""))

OPTIONS = ["Attribute1", "Attribute2", "Attribute3", "Attribute4"]

multi_choice = MultiChoice(value=["foo", "baz"], options=OPTIONS)

multi_choice.js_on_change("value", CustomJS(code="""

console.log('multi_choice: value=' + this.value, this.toString())

"""))

#GUI Left column

controls = [dropdown, spinner, toggle]

inputs = column(*controls, sizing_mode='fixed', height=250, width=350)

l1 = layout([[inputs, p1]], sizing_mode='fixed', height=600, width=150)

l2 = layout([[inputs, p2]], sizing_mode='fixed', height=600, width=150)

l3 = layout([[inputs, p3]], sizing_mode='fixed', height=600, width=150)

#Tab setup

tab1 = Panel(child=l1, title="Bar chart1")

tab2 = Panel(child=l2, title="Bar chart2")

tab3 = Panel(child=l3, title="Heat map")

tabs = Tabs(tabs=[tab1, tab2, tab3])

show(tabs)