kumify/mood/statistics.py

import holoviews as hv
import pandas as pd

from django.utils import timezone

from math import pi

from bokeh.models import HoverTool
from bokeh.plotting import figure
from bokeh.transform import cumsum
from bokeh.layouts import column
from holoviews.operation import timeseries
from dateutil.relativedelta import relativedelta

from .models import Status, Mood, StatusActivity


def moodstats(user):
    hv.extension("bokeh")

    tooltips = [("Date", "@date{%F %H:%M}"), ("Mood", "@name (@value)")]

    formatters = {"@date": "datetime"}

    hover = HoverTool(tooltips=tooltips, formatters=formatters)

    pointdict = {"date": [], "value": [], "color": [], "name": []}

    for status in Status.objects.filter(user=user):
        if status.mood:
            pointdict["date"].append(status.timestamp)
            pointdict["value"].append(status.mood.value)
            pointdict["color"].append(status.mood.color)
            pointdict["name"].append(status.mood.name)

    pointframe = pd.DataFrame.from_dict(pointdict)

    points = hv.Points(pointframe)

    points.opts(
        tools=[hover],
        color="color",
        cmap="Category20",
        line_color="black",
        size=25,
        width=600,
        height=400,
        show_grid=True,
    )

    pointtuples = [
        (pointdict["date"][i], pointdict["value"][i])
        for i in range(len(pointdict["date"]))
    ]

    line = hv.Curve(pointtuples)

    maxval = Mood.objects.filter(user=user).latest("value").value
    maxy = maxval + max(maxval * 0.1, 1)

    maxx = timezone.now().timestamp() * 1000
    minx = maxx - (60 * 60 * 24 * 7) * 1000

    output = points * line * timeseries.rolling(line, rolling_window=7)
    output.opts(ylim=(0, maxy), xlim=(minx, maxx))

    return output


def activitystats(user):
    output = {}

    for status in Status.objects.filter(user=user):
        for activity in status.activity_set:
            if not activity in output.keys():
                output[activity] = {
                    "alltime": 0,
                    "yearly": 0,
                    "monthly": 0,
                    "weekly": 0,
                }

            output[activity]["alltime"] += 1

            if status.timestamp > timezone.now() - relativedelta(years=1):
                output[activity]["yearly"] += 1

            if status.timestamp > timezone.now() - relativedelta(months=1):
                output[activity]["monthly"] += 1

            if status.timestamp > timezone.now() - relativedelta(weeks=1):
                output[activity]["weekly"] += 1

    return output


def moodpies(user):
    hv.extension("bokeh")

    maxdate = timezone.now()

    weekly_moods = Status.objects.filter(
        user=user,
        timestamp__lte=maxdate,
        timestamp__gte=maxdate - relativedelta(weeks=1),
    )
    monthly_moods = Status.objects.filter(
        user=user,
        timestamp__lte=maxdate,
        timestamp__gte=maxdate - relativedelta(months=1),
    )
    yearly_moods = Status.objects.filter(
        user=user,
        timestamp__lte=maxdate,
        timestamp__gte=maxdate - relativedelta(years=1),
    )

    weekly = dict()
    colors = []

    for mood in Mood.objects.filter(user=user):
        weekly[mood.name] = 0
        colors.append(mood.color)

    monthly, yearly = weekly.copy(), weekly.copy()

    for status in weekly_moods:
        if status.mood:
            weekly[status.mood.name] += 1

    for status in monthly_moods:
        if status.mood:
            monthly[status.mood.name] += 1

    for status in yearly_moods:
        if status.mood:
            yearly[status.mood.name] += 1

    weekly_data = (
        pd.Series(weekly).reset_index(name="value").rename(columns={"index": "mood"})
    )
    weekly_data["angle"] = weekly_data["value"] / weekly_data["value"].sum() * 2 * pi
    weekly_data["color"] = colors

    weekly_chart = figure(
        height=350,
        title="Weekly",
        toolbar_location=None,
        tools="hover",
        tooltips="@mood: @value",
    )
    weekly_chart.axis.visible = False

    weekly_chart.wedge(
        x=0,
        y=1,
        radius=0.4,
        start_angle=cumsum("angle", include_zero=True),
        end_angle=cumsum("angle"),
        line_color="white",
        fill_color="color",
        legend_label="mood",
        source=weekly_data,
    )

    monthly_data = (
        pd.Series(monthly).reset_index(name="value").rename(columns={"index": "mood"})
    )
    monthly_data["angle"] = monthly_data["value"] / monthly_data["value"].sum() * 2 * pi
    monthly_data["color"] = colors

    monthly_chart = figure(
        height=350,
        title="Monthly",
        toolbar_location=None,
        tools="hover",
        tooltips="@mood: @value",
    )
    monthly_chart.axis.visible = False

    monthly_chart.wedge(
        x=0,
        y=1,
        radius=0.4,
        start_angle=cumsum("angle", include_zero=True),
        end_angle=cumsum("angle"),
        line_color="white",
        fill_color="color",
        legend_label="mood",
        source=monthly_data,
    )

    yearly_data = (
        pd.Series(yearly).reset_index(name="value").rename(columns={"index": "mood"})
    )
    yearly_data["angle"] = yearly_data["value"] / yearly_data["value"].sum() * 2 * pi
    yearly_data["color"] = colors

    yearly_chart = figure(
        height=350,
        title="Yearly",
        toolbar_location=None,
        tools="hover",
        tooltips="@mood: @value",
    )
    yearly_chart.axis.visible = False

    yearly_chart.wedge(
        x=0,
        y=1,
        radius=0.4,
        start_angle=cumsum("angle", include_zero=True),
        end_angle=cumsum("angle"),
        line_color="white",
        fill_color="color",
        legend_label="mood",
        source=yearly_data,
    )

    return column(weekly_chart, monthly_chart, yearly_chart)


def activitymood(activity):
    hv.extension("bokeh")

    tooltips = [("Date", "@date{%F %H:%M}"), ("Mood", "@name (@value)")]

    formatters = {"@date": "datetime"}

    hover = HoverTool(tooltips=tooltips, formatters=formatters)

    pointdict = {"date": [], "value": [], "color": [], "name": []}

    for statusactivity in StatusActivity.objects.filter(activity=activity):
        if statusactivity.status.mood:
            pointdict["date"].append(statusactivity.status.timestamp)
            pointdict["value"].append(statusactivity.status.mood.value)
            pointdict["color"].append(statusactivity.status.mood.color)
            pointdict["name"].append(statusactivity.status.mood.name)

    pointframe = pd.DataFrame.from_dict(pointdict)

    points = hv.Points(pointframe)

    points.opts(
        tools=[hover],
        color="color",
        cmap="Category20",
        line_color="black",
        size=25,
        width=600,
        height=400,
        show_grid=True,
    )

    pointtuples = [
        (pointdict["date"][i], pointdict["value"][i])
        for i in range(len(pointdict["date"]))
    ]

    line = hv.Curve(pointtuples)

    maxval = Mood.objects.filter(user=activity.user).latest("value").value
    maxy = maxval + max(maxval * 0.1, 1)

    maxx = timezone.now().timestamp() * 1000
    minx = maxx - (60 * 60 * 24 * 7) * 1000

    output = points * line * timeseries.rolling(line, rolling_window=7)
    output.opts(ylim=(0, maxy), xlim=(minx, maxx))

    return output


def activitypies(activity):
    hv.extension("bokeh")

    maxdate = timezone.now()

    sa = StatusActivity.objects.filter(activity=activity)

    weekly = dict()
    colors = []

    for mood in Mood.objects.filter(user=activity.user):
        weekly[mood.name] = 0
        colors.append(mood.color)

    monthly, yearly = weekly.copy(), weekly.copy()

    for single in sa:
        if single.status.mood:
            if single.status.timestamp > timezone.now() - relativedelta(weeks=1):
                weekly[single.status.mood.name] += 1
            if single.status.timestamp > timezone.now() - relativedelta(months=1):
                monthly[single.status.mood.name] += 1
            if single.status.timestamp > timezone.now() - relativedelta(years=1):
                yearly[single.status.mood.name] += 1

    weekly_data = (
        pd.Series(weekly).reset_index(name="value").rename(columns={"index": "mood"})
    )
    weekly_data["angle"] = weekly_data["value"] / weekly_data["value"].sum() * 2 * pi
    weekly_data["color"] = colors

    weekly_chart = figure(
        height=350,
        title="Weekly",
        toolbar_location=None,
        tools="hover",
        tooltips="@mood: @value",
    )
    weekly_chart.axis.visible = False

    weekly_chart.wedge(
        x=0,
        y=1,
        radius=0.4,
        start_angle=cumsum("angle", include_zero=True),
        end_angle=cumsum("angle"),
        line_color="white",
        fill_color="color",
        legend_label="mood",
        source=weekly_data,
    )

    monthly_data = (
        pd.Series(monthly).reset_index(name="value").rename(columns={"index": "mood"})
    )
    monthly_data["angle"] = monthly_data["value"] / monthly_data["value"].sum() * 2 * pi
    monthly_data["color"] = colors

    monthly_chart = figure(
        height=350,
        title="Monthly",
        toolbar_location=None,
        tools="hover",
        tooltips="@mood: @value",
    )
    monthly_chart.axis.visible = False

    monthly_chart.wedge(
        x=0,
        y=1,
        radius=0.4,
        start_angle=cumsum("angle", include_zero=True),
        end_angle=cumsum("angle"),
        line_color="white",
        fill_color="color",
        legend_label="mood",
        source=monthly_data,
    )

    yearly_data = (
        pd.Series(yearly).reset_index(name="value").rename(columns={"index": "mood"})
    )
    yearly_data["angle"] = yearly_data["value"] / yearly_data["value"].sum() * 2 * pi
    yearly_data["color"] = colors

    yearly_chart = figure(
        height=350,
        title="Yearly",
        toolbar_location=None,
        tools="hover",
        tooltips="@mood: @value",
    )
    yearly_chart.axis.visible = False

    yearly_chart.wedge(
        x=0,
        y=1,
        radius=0.4,
        start_angle=cumsum("angle", include_zero=True),
        end_angle=cumsum("angle"),
        line_color="white",
        fill_color="color",
        legend_label="mood",
        source=yearly_data,
    )

    return column(weekly_chart, monthly_chart, yearly_chart)