#+TITLE:  Courte "analyse" de l'évolution de la température 
#+AUTHOR: Miguel Arpa Perozo
#+DATE:   <2020-10-29 jeu.>
#+LANGUAGE: fr
# #+PROPERTY: header-args :eval never-export

#+HTML_HEAD: <link rel="stylesheet" type="text/css" href="http://www.pirilampo.org/styles/readtheorg/css/htmlize.css"/>
#+HTML_HEAD: <link rel="stylesheet" type="text/css" href="http://www.pirilampo.org/styles/readtheorg/css/readtheorg.css"/>
#+HTML_HEAD: <script src="https://ajax.googleapis.com/ajax/libs/jquery/2.1.3/jquery.min.js"></script>
#+HTML_HEAD: <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.4/js/bootstrap.min.js"></script>
#+HTML_HEAD: <script type="text/javascript" src="http://www.pirilampo.org/styles/lib/js/jquery.stickytableheaders.js"></script>
#+HTML_HEAD: <script type="text/javascript" src="http://www.pirilampo.org/styles/readtheorg/js/readtheorg.js"></script>

* Téléchargement des données 
  Le site méteo france met en accès librement des données météorologique gratuitement : https://donneespubliques.meteofrance.fr/. 
  
  Voici un petit script =bash= qui permet de télécharger automatiquement toutes les données depuis 1996 jusqu'à 2020 : 

  #+begin_src bash :exports code
  #!/usr/bin/bash
  baseLink="https://donneespubliques.meteofrance.fr/donnees_libres/Txt/Synop/Archive/synop."
  endLink=".csv.gz"
  outputDest="./download"
  for i in `seq 1996 2020`; do
    for j in `seq -f "%02g" 12`; do
	link="${baseLink}${i}${j}${endLink}"
	wget $link -P $outputDest 
    done
  done
  #+end_src
  Toutes les données sont ensuite extraites et disponibles en format =.csv= dans le dossier =./data=. 

* Conventions et fonctions utiles 
  Le script suivant contient : 
  - Des constantes qui permettent de faire l'interface avec les conventions utilisées par les stations météorologiques.
  - Variables contenant les données d'intérêt. 
    
  #+begin_src python :results output :session :exports code
    import numpy as np
    from scipy import io as spio 
    from matplotlib import pyplot as plt

    #Kelvin to celsius
    k2deg = 273.15

    #Station's ids 
    stationId = {
	 "strasbourg": 7190,
	 "nancy":      7181,
	 "mulhouse":   7299} 

    #Column index of the different values 
    columnIndx = {
	 "station":     0,
	 "date":        1,
	 "pressure":    2,
	 "temperature": 7,
	 "humidity":    9
    }

    # List containing all the raw data of the day of interest by year 
    usefullData   = []
    dataPath  = "./data/synop."
    dataMonth = "10.csv"
    month = 10
    day = 21
  #+end_src

  #+RESULTS:

  - Fonction qui permet d'obtenir les données pour un jour précis.
  #+begin_src python :results output :session :exports both
    def getDayData(year,month,day,data):
        """ Returns only the information for a specific day.
        The data is encoded as a number with the format : 
        AAAAMMDDHHMISS"""

        if month <10 and day <10:
            dayString = str(year) + '0' + str(month) + '0' + str(day) # year + month + day
        elif month <10:
            dayString = str(year) + '0' + str(month) + str(day) # year + month + day
        elif  day <10:
            dayString = str(year) + str(month) + '0' + str(day) # year + month + day
        else:
            dayString = str(year) + str(month) + str(day) # year + month + day

        dateInfo = 4+2+2
        filteredData = np.array([])
        for row in data:
            timeStamp = str(row[1]); #location of the timestamp
            date = timeStamp[0:dateInfo]
            if date == dayString:
                filteredData = np.concatenate((filteredData, row),axis=0)
            foundRows = int(filteredData.shape[0] / data.shape[1])
            filteredData =  filteredData.reshape((foundRows,data.shape[1]))
        return filteredData
  #+end_src

  #+RESULTS:

* Traitement des données  
  #+begin_src python :results output :session :exports both
    for year in range(1996,2020):
	fileName = dataPath+ str(year) + dataMonth
	#print(fileName)
	rawData = np.genfromtxt(fileName,delimiter=";",skip_header=1)
	strasbourg = rawData[ rawData[:,columnIndx["station"]] == stationId["strasbourg"]]
	temperature = strasbourg[:,columnIndx["temperature"]] - k2deg; 
	usefullData.append(temperature)



    minTemp    = np.zeros(len(usefullData))
    maxTemp    = np.zeros(len(usefullData))
    meanTemp   = np.zeros(len(usefullData))
    medianTemp = np.zeros(len(usefullData))

    years   = np.arange(1996,2020)

    cnt = 0
    for temp in usefullData:
	minTemp[cnt]    = np.min(temp)
	maxTemp[cnt]    = np.max(temp)
	meanTemp[cnt]   = np.mean(temp)
	medianTemp[cnt] = np.median(temp)
	cnt += 1
  #+end_src

  #+RESULTS:

* Résultats 
#+begin_src python :results output file :session :var matplot_lib_filename="results.png" :exports both
  plt.scatter(years,minTemp,label="min")
  plt.scatter(years,maxTemp,label="max")
  plt.scatter(years,meanTemp,label="mean")
  plt.scatter(years,medianTemp,label="median")
  plt.title('Analyse de le température par an'); 
  plt.xlabel('An'); 
  plt.ylabel('Température [°C]'); 
  plt.legend(); 
  plt.savefig(matplot_lib_filename)
  print(matplot_lib_filename)
#+end_src

#+RESULTS:
[[file:results.png]]