jku-ba-react/src/database/dbAnalyser.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Fri Nov 27 23:48:19 2020

@author: bernadetteelena
"""
import os
import json
import pandas as pd
import matplotlib.pyplot as plt

colorLookup = {
    "google": "r",
    "leaflet": "b",
    "mapbox_feature": "g",
    "mapbox_marker": "y",
    "mapgl": "m",
    "pigeons": "c"
}


def readFile():
    filePath = os.path.dirname(os.path.abspath(__file__)) + "/time.json"
    with open(filePath) as file:
        timeList = json.load(file)

    timeDf = pd.DataFrame(timeList)
    print("number of tests:", timeDf.shape[0])
    timeDf["markers"] = pd.to_numeric(timeDf["markers"], errors="raise")
    print(timeDf.value_counts(subset=["id"]))
    print(timeDf.value_counts(subset=["id", "type"]))
    return timeDf


def timeMaps(df, plotTitle="maps.png"):
    fig, ax = plt.subplots(figsize=(20,12))
    plt.yticks(fontsize=15)
    plt.xticks(fontsize=15)
    for mapId, mapDf in df.groupby(["id"]):
        if mapId != "mapbox":
            ax.scatter(mapDf["markers"], mapDf["time"], color=colorLookup.get(mapId))
            mapAvgDf = mapDf.groupby(["markers"]).mean().reset_index()
            ax.plot(mapAvgDf["markers"], mapAvgDf["time"], label=mapId, color=colorLookup.get(mapId))
        else:
            for markerType, typeDf in mapDf.groupby(["type"]):
                mapTypeId = "mapbox_" + markerType
                ax.scatter(typeDf["markers"], typeDf["time"], color=colorLookup.get(mapTypeId))
                typeMapAvgDf = typeDf.groupby(["markers"]).mean().reset_index()
                ax.plot(typeMapAvgDf["markers"], typeMapAvgDf["time"], label=mapTypeId, color=colorLookup.get(mapTypeId))

    ax.legend(title="Marker Type", fontsize=17, title_fontsize=20)
    ax.set_xlabel("Number of Markers", fontsize=20)
    ax.set_ylabel("Time (ms)", fontsize=20)
    ax.set_title("Map Libraries - Marker", fontsize=25)
    folder = os.path.dirname(os.path.abspath(__file__)) + "/plots/"
    plt.savefig(folder + plotTitle, bbox_inches="tight")


def mapboxTypes(df, plotTitle="mapbox.png"):
    mapboxDf = df.dropna(subset=["type"])
    fig, ax = plt.subplots(figsize=(20,12))
    plt.yticks(fontsize=15)
    plt.xticks(fontsize=15)
    for markerType, typeDf in mapboxDf.groupby(["type"]):
        mapId = "mapbox_" + markerType
        ax.scatter(typeDf["markers"], typeDf["time"], label=mapId, color=colorLookup.get(mapId))

        typeAvgDf = typeDf.groupby(["markers"]).mean().reset_index()
        ax.plot(typeAvgDf["markers"], typeAvgDf["time"], color=colorLookup.get(mapId))

    ax.legend(title="Marker Type", fontsize=17, title_fontsize=20)
    ax.set_xlabel("Number of Markers", fontsize=20)
    ax.set_ylabel("Time (ms)", fontsize=20)
    ax.set_title("Mapbox: Markers and Features", fontsize=25)
    folder = os.path.dirname(os.path.abspath(__file__)) + "/plots/"
    plt.savefig(folder + plotTitle, bbox_inches="tight")


def smallValues(df):
    df = df.loc[df["markers"] < 300]
    timeMaps(df, "mapsSmall.png")
    mapboxTypes(df, "mapboxSmall.png")


def avgPerMarkerNo(df, markerNo):
    print(f"\naverage time for {markerNo} markers")
    print("number of measurements")
    df = df.loc[df["markers"] == markerNo]
    resDf = pd.DataFrame(columns=["id", "timeMean", "rating"])
    for mapId, mapDf in df.groupby(["id"]):
        if mapId != "mapbox":
            print(mapId, mapDf.shape[0])
            resDf = resDf.append({"id": mapId, "timeMean": mapDf["time"].mean()}, ignore_index=True)
        else:
            for markerType, typeDf in mapDf.groupby(["type"]):
                print(markerType, typeDf.shape[0])
                resDf = resDf.append({"id": "mapbox_" + markerType, "timeMean": typeDf["time"].mean()}, ignore_index=True)

    resDf = resDf.assign(rating=lambda x: (round(4-((4-0)/(max(resDf["timeMean"])-min(resDf["timeMean"]))*(x["timeMean"]-max(resDf["timeMean"]))+4))))
    print(resDf)



df = readFile()
avgPerMarkerNo(df, 100)
avgPerMarkerNo(df, 1000)
avgPerMarkerNo(df, 10000)
timeMaps(df)
mapboxTypes(df)
smallValues(df)