{ "cells": [ { "cell_type": "markdown", "metadata": { "hideCode": false, "hidePrompt": false }, "source": [ "# Concentration de CO2 dans l'atmosphère depuis 1958\n", "## Importation des bibliothèques" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "hideCode": false, "hidePrompt": false }, "outputs": [], "source": [ "%matplotlib inline\n", "import matplotlib.pyplot as plt\n", "import pandas as pd\n", "import isoweek\n", "import os\n", "import urllib.request\n", "import numpy as np\n", "import sys\n", "#Installation de la bibliothèque lmfit (si jamais l'utilisateur ne l'a pas, autrement à mettre en commentaire)\n", "!{sys.executable} -m pip install lmfit \n", "import lmfit" ] }, { "cell_type": "markdown", "metadata": { "hideCode": false, "hidePrompt": false }, "source": [ "Les donnéessur la concentration de CO$_2$ dans l'atmosphère sont disponibles du site Web de l'[institut Scripps](https://scrippsco2.ucsd.edu/data/atmospheric_co2/primary_mlo_co2_record.html). Nous les récupérons sous forme d'un fichier en format CSV dont chaque ligne correspond à un mois de la période demandée. Nous téléchargeons toujours le jeu de données complet, qui commence en 1958 et se termine au 1er janvier 2020." ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "hideCode": false, "hidePrompt": false }, "outputs": [], "source": [ "data_url = \"https://scrippsco2.ucsd.edu/assets/data/atmospheric/stations/in_situ_co2/monthly/monthly_in_situ_co2_mlo.csv\"" ] }, { "cell_type": "markdown", "metadata": { "hideCode": false, "hidePrompt": false }, "source": [ "Importation des données : Nous faisons une copie locale du fichier de données au cas où il y a modification du serveur du site de l'institut Scripps. Il est plus sûr de faire une copie du fichier en local au cas où le format/url change, pour travailler sur le même jeu de données et être reproductible." ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "hideCode": false, "hidePrompt": false }, "outputs": [], "source": [ "data_file = \"Concentration_CO2.csv\"\n", "\n", "if not os.path.exists(data_file):\n", " urllib.request.urlretrieve(data_url, data_file)" ] }, { "cell_type": "markdown", "metadata": { "hideCode": false, "hidePrompt": false }, "source": [ "Voici l'explication des colonnes données [sur le site d'origine](https://scrippsco2.ucsd.edu/data/atmospheric_co2/primary_mlo_co2_record.html):\n", "\n", "| Nom de colonne | Libellé de colonne |\n", "|----------------|-----------------------------------------------------------------------------------------------------------------------------------|\n", "| Yr | Année de la mesure |\n", "| Mn | Mois de la mesure |\n", "| Date | Date de la mesure au format Excel |\n", "| Date | Date de la mesure sous un autre format\n", "| CO2 | Mesure de la concentration en CO2 en micro-mol CO2 per mol (ppm) |\n", "| seasonally | Même mesure mais avec soustraction de l'influence saisonniaire. |\n", "| fit | Ajustement de la courbe de la colonne 5 par un modèle de fonction cubique avec 4 harmoniques et un gain linéaire. |\n", "| seasonally | Ajustement de la courbe de la colonne 4 par un modèle de fonction cubique avec 4 harmoniques et un gain linéaire. |\n", "| CO2 | Copie de la colonne 5 avec les valeurs manquantes remplacées par les valeurs de la colonne 7. |\n", "| seasonally | Copie de la colonne 6 avec les valeurs manquantes remplacées par les valeurs de la colonne 8 |\n", " |\n", "\n", "La première ligne du fichier CSV est un commentaire, que nous ignorons en précisant `skiprows=54`.\n", "Les **valeurs manquantes** sont dénotées **-99.99**.\n", "\n", "## Mise en forme des données" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "hideCode": false, "hidePrompt": false }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
YrMnDateDateCO2seasonallyfitseasonallyCO2seasonally
0adjustedadjusted fitfilledadjusted filled
1Excel[ppm][ppm][ppm][ppm][ppm][ppm]
2195801212001958.0411-99.99-99.99-99.99-99.99-99.99-99.99
3195802212311958.1260-99.99-99.99-99.99-99.99-99.99-99.99
4195803212591958.2027315.70314.44316.19314.91315.70314.44
5195804212901958.2877317.45315.16317.30314.99317.45315.16
6195805213201958.3699317.51314.71317.86315.06317.51314.71
7195806213511958.4548-99.99-99.99317.24315.14317.24315.14
8195807213811958.5370315.86315.19315.86315.22315.86315.19
9195808214121958.6219314.93316.19314.00315.29314.93316.19
10195809214431958.7068313.21316.08312.46315.35313.21316.08
11195810214731958.7890-99.99-99.99312.44315.40312.44315.40
12195811215041958.8740313.33315.20313.62315.46313.33315.20
13195812215341958.9562314.67315.43314.77315.51314.67315.43
14195901215651959.0411315.58315.54315.62315.57315.58315.54
15195902215961959.1260316.49315.86316.27315.63316.49315.86
16195903216241959.2027316.65315.38316.98315.69316.65315.38
17195904216551959.2877317.72315.42318.09315.77317.72315.42
18195905216851959.3699318.29315.49318.65315.85318.29315.49
19195906217161959.4548318.15316.03318.04315.94318.15316.03
20195907217461959.5370316.54315.86316.67316.03316.54315.86
21195908217771959.6219314.80316.06314.83316.12314.80316.06
22195909218081959.7068313.84316.72313.32316.22313.84316.72
23195910218381959.7890313.33316.32313.33316.30313.33316.32
24195911218691959.8740314.81316.68314.54316.39314.81316.68
25195912218991959.9562315.58316.35315.72316.47315.58316.35
26196001219301960.0410316.43316.39316.61316.56316.43316.39
27196002219611960.1257316.98316.35317.27316.64316.98316.35
28196003219901960.2049317.58316.28318.03316.71317.58316.28
29196004220211960.2896319.03316.70319.14316.79319.03316.70
.................................
728201807432962018.5370408.90408.08409.43408.65408.90408.08
729201808433272018.6219407.10408.63407.33408.91407.10408.63
730201809433582018.7068405.59409.09405.66409.18405.59409.09
731201810433882018.7890405.99409.62405.83409.44405.99409.62
732201811434192018.8740408.12410.39407.47409.72408.12410.39
733201812434492018.9562409.23410.16409.07409.97409.23410.16
734201901434802019.0411410.92410.87410.29410.23410.92410.87
735201902435112019.1260411.66410.90411.24410.47411.66410.90
736201903435392019.2027412.00410.45412.25410.68412.00410.45
737201904435702019.2877413.52410.72413.73410.91413.52410.72
738201905436002019.3699414.83411.42414.54411.13414.83411.42
739201906436312019.4548413.96411.38413.91411.35413.96411.38
740201907436612019.5370411.85411.03412.36411.57411.85411.03
741201908436922019.6219410.08411.62410.23411.81410.08411.62
742201909437232019.7068408.55412.06408.52412.05408.55412.06
743201910437532019.7890408.43412.07408.67412.29408.43412.07
744201911437842019.8740410.28412.56410.29412.54410.28412.56
745201912438142019.9562411.85412.78411.88412.79411.85412.78
746202001438452020.0410413.37413.33413.11413.05413.37413.33
747202002438762020.1257-99.99-99.99-99.99-99.99-99.99-99.99
748202003439052020.2049-99.99-99.99-99.99-99.99-99.99-99.99
749202004439362020.2896-99.99-99.99-99.99-99.99-99.99-99.99
750202005439662020.3716-99.99-99.99-99.99-99.99-99.99-99.99
751202006439972020.4563-99.99-99.99-99.99-99.99-99.99-99.99
752202007440272020.5383-99.99-99.99-99.99-99.99-99.99-99.99
753202008440582020.6230-99.99-99.99-99.99-99.99-99.99-99.99
754202009440892020.7077-99.99-99.99-99.99-99.99-99.99-99.99
755202010441192020.7896-99.99-99.99-99.99-99.99-99.99-99.99
756202011441502020.8743-99.99-99.99-99.99-99.99-99.99-99.99
757202012441802020.9563-99.99-99.99-99.99-99.99-99.99-99.99
\n", "

758 rows × 10 columns

\n", "
" ], "text/plain": [ " Yr Mn Date Date CO2 seasonally fit \\\n", "0 adjusted \n", "1 Excel [ppm] [ppm] [ppm] \n", "2 1958 01 21200 1958.0411 -99.99 -99.99 -99.99 \n", "3 1958 02 21231 1958.1260 -99.99 -99.99 -99.99 \n", "4 1958 03 21259 1958.2027 315.70 314.44 316.19 \n", "5 1958 04 21290 1958.2877 317.45 315.16 317.30 \n", "6 1958 05 21320 1958.3699 317.51 314.71 317.86 \n", "7 1958 06 21351 1958.4548 -99.99 -99.99 317.24 \n", "8 1958 07 21381 1958.5370 315.86 315.19 315.86 \n", "9 1958 08 21412 1958.6219 314.93 316.19 314.00 \n", "10 1958 09 21443 1958.7068 313.21 316.08 312.46 \n", "11 1958 10 21473 1958.7890 -99.99 -99.99 312.44 \n", "12 1958 11 21504 1958.8740 313.33 315.20 313.62 \n", "13 1958 12 21534 1958.9562 314.67 315.43 314.77 \n", "14 1959 01 21565 1959.0411 315.58 315.54 315.62 \n", "15 1959 02 21596 1959.1260 316.49 315.86 316.27 \n", "16 1959 03 21624 1959.2027 316.65 315.38 316.98 \n", "17 1959 04 21655 1959.2877 317.72 315.42 318.09 \n", "18 1959 05 21685 1959.3699 318.29 315.49 318.65 \n", "19 1959 06 21716 1959.4548 318.15 316.03 318.04 \n", "20 1959 07 21746 1959.5370 316.54 315.86 316.67 \n", "21 1959 08 21777 1959.6219 314.80 316.06 314.83 \n", "22 1959 09 21808 1959.7068 313.84 316.72 313.32 \n", "23 1959 10 21838 1959.7890 313.33 316.32 313.33 \n", "24 1959 11 21869 1959.8740 314.81 316.68 314.54 \n", "25 1959 12 21899 1959.9562 315.58 316.35 315.72 \n", "26 1960 01 21930 1960.0410 316.43 316.39 316.61 \n", "27 1960 02 21961 1960.1257 316.98 316.35 317.27 \n", "28 1960 03 21990 1960.2049 317.58 316.28 318.03 \n", "29 1960 04 22021 1960.2896 319.03 316.70 319.14 \n", ".. ... ... ... ... ... ... ... \n", "728 2018 07 43296 2018.5370 408.90 408.08 409.43 \n", "729 2018 08 43327 2018.6219 407.10 408.63 407.33 \n", "730 2018 09 43358 2018.7068 405.59 409.09 405.66 \n", "731 2018 10 43388 2018.7890 405.99 409.62 405.83 \n", "732 2018 11 43419 2018.8740 408.12 410.39 407.47 \n", "733 2018 12 43449 2018.9562 409.23 410.16 409.07 \n", "734 2019 01 43480 2019.0411 410.92 410.87 410.29 \n", "735 2019 02 43511 2019.1260 411.66 410.90 411.24 \n", "736 2019 03 43539 2019.2027 412.00 410.45 412.25 \n", "737 2019 04 43570 2019.2877 413.52 410.72 413.73 \n", "738 2019 05 43600 2019.3699 414.83 411.42 414.54 \n", "739 2019 06 43631 2019.4548 413.96 411.38 413.91 \n", "740 2019 07 43661 2019.5370 411.85 411.03 412.36 \n", "741 2019 08 43692 2019.6219 410.08 411.62 410.23 \n", "742 2019 09 43723 2019.7068 408.55 412.06 408.52 \n", "743 2019 10 43753 2019.7890 408.43 412.07 408.67 \n", "744 2019 11 43784 2019.8740 410.28 412.56 410.29 \n", "745 2019 12 43814 2019.9562 411.85 412.78 411.88 \n", "746 2020 01 43845 2020.0410 413.37 413.33 413.11 \n", "747 2020 02 43876 2020.1257 -99.99 -99.99 -99.99 \n", "748 2020 03 43905 2020.2049 -99.99 -99.99 -99.99 \n", "749 2020 04 43936 2020.2896 -99.99 -99.99 -99.99 \n", "750 2020 05 43966 2020.3716 -99.99 -99.99 -99.99 \n", "751 2020 06 43997 2020.4563 -99.99 -99.99 -99.99 \n", "752 2020 07 44027 2020.5383 -99.99 -99.99 -99.99 \n", "753 2020 08 44058 2020.6230 -99.99 -99.99 -99.99 \n", "754 2020 09 44089 2020.7077 -99.99 -99.99 -99.99 \n", "755 2020 10 44119 2020.7896 -99.99 -99.99 -99.99 \n", "756 2020 11 44150 2020.8743 -99.99 -99.99 -99.99 \n", "757 2020 12 44180 2020.9563 -99.99 -99.99 -99.99 \n", "\n", " seasonally CO2 seasonally \n", "0 adjusted fit filled adjusted filled \n", "1 [ppm] [ppm] [ppm] \n", "2 -99.99 -99.99 -99.99 \n", "3 -99.99 -99.99 -99.99 \n", "4 314.91 315.70 314.44 \n", "5 314.99 317.45 315.16 \n", "6 315.06 317.51 314.71 \n", "7 315.14 317.24 315.14 \n", "8 315.22 315.86 315.19 \n", "9 315.29 314.93 316.19 \n", "10 315.35 313.21 316.08 \n", "11 315.40 312.44 315.40 \n", "12 315.46 313.33 315.20 \n", "13 315.51 314.67 315.43 \n", "14 315.57 315.58 315.54 \n", "15 315.63 316.49 315.86 \n", "16 315.69 316.65 315.38 \n", "17 315.77 317.72 315.42 \n", "18 315.85 318.29 315.49 \n", "19 315.94 318.15 316.03 \n", "20 316.03 316.54 315.86 \n", "21 316.12 314.80 316.06 \n", "22 316.22 313.84 316.72 \n", "23 316.30 313.33 316.32 \n", "24 316.39 314.81 316.68 \n", "25 316.47 315.58 316.35 \n", "26 316.56 316.43 316.39 \n", "27 316.64 316.98 316.35 \n", "28 316.71 317.58 316.28 \n", "29 316.79 319.03 316.70 \n", ".. ... ... ... \n", "728 408.65 408.90 408.08 \n", "729 408.91 407.10 408.63 \n", "730 409.18 405.59 409.09 \n", "731 409.44 405.99 409.62 \n", "732 409.72 408.12 410.39 \n", "733 409.97 409.23 410.16 \n", "734 410.23 410.92 410.87 \n", "735 410.47 411.66 410.90 \n", "736 410.68 412.00 410.45 \n", "737 410.91 413.52 410.72 \n", "738 411.13 414.83 411.42 \n", "739 411.35 413.96 411.38 \n", "740 411.57 411.85 411.03 \n", "741 411.81 410.08 411.62 \n", "742 412.05 408.55 412.06 \n", "743 412.29 408.43 412.07 \n", "744 412.54 410.28 412.56 \n", "745 412.79 411.85 412.78 \n", "746 413.05 413.37 413.33 \n", "747 -99.99 -99.99 -99.99 \n", "748 -99.99 -99.99 -99.99 \n", "749 -99.99 -99.99 -99.99 \n", "750 -99.99 -99.99 -99.99 \n", "751 -99.99 -99.99 -99.99 \n", "752 -99.99 -99.99 -99.99 \n", "753 -99.99 -99.99 -99.99 \n", "754 -99.99 -99.99 -99.99 \n", "755 -99.99 -99.99 -99.99 \n", "756 -99.99 -99.99 -99.99 \n", "757 -99.99 -99.99 -99.99 \n", "\n", "[758 rows x 10 columns]" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "raw_data = pd.read_csv(data_url, skiprows=54)\n", "raw_data" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "On peut remarquer que les deux premières lignes sont des commentaires pour se repérer dans les colonnes, elles nous sont inutiles pour le traitement. De même que les valeurs manquantes dénotées -99.99, il nous faut les supprimer.\n", "Commençons par récupérer les colonnes qui nous intéressent, d'en supprimer les 2 premières lignes et de les convertir en nombre :" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [], "source": [ "Dates=raw_data[' Date'][2:].values\n", "data=raw_data[' CO2'][2:].values\n", "for i in range(len(data)):\n", " data[i]=float(data[i])\n", " Dates[i]=float(Dates[i])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Cherchons maintenant à en supprimer les valeurs manquantes :" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [], "source": [ "def Reduce(Dates,data):\n", " reduced_dates=list()\n", " reduced_data=list()\n", " for i in range(len(Dates)):\n", " if data[i]!=-99.99:\n", " reduced_data.append(data[i])\n", " reduced_dates.append(Dates[i])\n", " return reduced_data, reduced_dates\n", "reduced_data, reduced_dates=Reduce(Dates,data)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Regardons maintenant l'allure de la courbe :" ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "hideCode": true }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "plt.plot(reduced_dates,reduced_data,label='Concentration$_{CO_{2}}$=f(temps)')\n", "plt.legend()\n", "plt.grid()\n", "plt.xlabel('Dates')\n", "plt.ylabel('Concentration en CO$_{2}$ (ppm)')\n", "plt.title('Représentation des données réduites');" ] }, { "cell_type": "markdown", "metadata": { "hideCode": true }, "source": [ "On observe que nous avons une oscillation périodique avec une tendance à la croître en fonction du temps. \n", "\n", "## Caractérisation des oscillations et de la tendance\n", "\n", "1. Recherche d'un modèle général de la courbe\n", "\n", "Dans le header du fichier de données, on nous dis que les données peuvent être ajustées par une fonction cubique et une somme d'harmonique. La fonction cubique représente la tendance de la courbe à augmenter au cours des ans tandis que la somme des harmoniques représente les oscillations de la courbe.\n", "\n", "Nous définissons donc une fonction dont nous ajusterons ses paramètres à l'aide de la bibliothèque _lmfit_ par la méthode de [Levenberg-Marquardt](https://fr.wikipedia.org/wiki/Algorithme_de_Levenberg-Marquardt). La fonction proposée est la suivante :\n", "\n", "$Model=a+b*(x-x_{0})^{c}+d*\\frac{cos(2\\pi*f*x)}{1!}+e*\\frac{cos(2\\pi*3f*x)}{3!}+g*\\frac{cos(2\\pi*5f*x)}{5!}+h*\\frac{cos(2\\pi*7f*x)}{7!}$\n", "\n", "Nous choisissons ici d'exprimer la fonction cubique comme $a+b*(x-x_{0})^{c}$, avec _a_ l'ordonnée à l'origine, _b_ une constante de multiplication, _$x_{0}$_ une constante pour décaler la courbe dans le temps et _c_ une constante proche de 3 (pour avoir une fonction cubique).\n", "\n", "Concernant le reste du modèle, cela représente le développement en série d'une fonction triangle jusqu'à la 4$^{e}$ harmonique. Comme ce développement comprend uniquement les harmoniques impaires, les harmoniques pertinentes sont 1,3,5 et 7. Le choix de la fonction triangle se justifie par l'allure de la courbe \"à l'oeil\", ce qui peut-être discutable bien évidemment. \n", "\n", "Par rapport aux noms des variables, celle qui nous intéresse le plus est la fréquence _f_, les autres termes _d,e,g_ et _h_ sont des constantes de multiplication pour l'ajustement de la courbe.\n", "\n", "Les variables sont ensuite initialisées dans un vecteur w contenant des valeurs proches de celles attendues ou contenant des 1 pour celles dont on ne connaît absolument pas la valeur. On fait de même varier les variables entre des valeurs possibles." ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "scrolled": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[[Model]]\n", " Model(FonctionModel)\n", "[[Fit Statistics]]\n", " # fitting method = leastsq\n", " # function evals = 349\n", " # data points = 743\n", " # variables = 9\n", " chi-square = 645.056908\n", " reduced chi-square = 0.87882413\n", " Akaike info crit = -87.0286239\n", " Bayesian info crit = -45.5323595\n", "[[Variables]]\n", " t0: 1870.91642 +/- 25.0616894 (1.34%) (init = 1951)\n", " A: 2.84439867 +/- 0.04861780 (1.71%) (init = 1)\n", " B: 1.00000043 +/- 0.31892352 (31.89%) (init = 1)\n", " C: 0.64346148 +/- 1.25795087 (195.50%) (init = 1)\n", " D: 3.6390e-06 +/- 2.46794175 (67818852.73%) (init = 1)\n", " E: 3.8765e-05 +/- 9.9882e-05 (257.66%) (init = 1)\n", " F: 289.430736 +/- 4.19785528 (1.45%) (init = 300)\n", " G: 2.99059870 +/- 0.42153584 (14.10%) (init = 3)\n", " a: 0.99984556 +/- 1.3619e-06 (0.00%) (init = 1)\n", "[[Correlations]] (unreported correlations are < 0.100)\n", " C(E, G) = -1.000\n", " C(t0, E) = 0.999\n", " C(t0, G) = -0.999\n", " C(t0, F) = 0.996\n", " C(E, F) = 0.991\n", " C(F, G) = -0.990\n", " C(C, D) = -0.209\n" ] } ], "source": [ "#Définition du modèle\n", "def FonctionModel(t,t0,A,B,C,D,E,F,G,a):\n", " return F+E*(t-t0)**G+A*np.cos(2*np.pi*a*t)+B*np.cos(2*np.pi*3*a*t)/9+C*np.cos(2*np.pi*5*a*t)/25+D*np.cos(2*np.pi*7*a*t)/49\n", " \n", "gmodel = lmfit.Model(FonctionModel)\n", "\n", "#Définition des paramètres d'ajustement\n", "params=[]\n", "params = lmfit.Parameters()\n", "w=[1951,1,1,1,1,1,300,3,1]\n", "params.add('t0', value=w[0], vary=True, min=0, max=1960)\n", "params.add('A', value=w[1], vary=True, min=1, max=10000)\n", "params.add('B', value=w[2], vary=True, min=1, max=10000)\n", "params.add('C', value=w[3], vary=True, min=0, max=40000)\n", "params.add('D', value=w[4], vary=True, min=0, max=40000)\n", "params.add('E', value=w[5], vary=True, min=0, max=40000)\n", "params.add('F', value=w[6], vary=True, min=0, max=40000)\n", "params.add('G', value=w[7], vary=True, min=2.99, max=3.01)\n", "params.add('a', value=w[8], vary=True, min=0.5, max=10)\n", "\n", "#Ajustement\n", "result=gmodel.fit(reduced_data,params,t=reduced_dates, scale_covar=True, fit_kws={'ftol': 1e-10, 'xtol': 1e-10, 'gtol': 1e-10})\n", "print(result.fit_report())" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "On remarque dans un premier temps que le $\\chi^{2}$ est élevé, mais néanmoins regardons l'allure de la courbe ajustée." ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "hideCode": true, "scrolled": true }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "plt.plot(reduced_dates,reduced_data,label='Données réduites')\n", "plt.errorbar(reduced_dates,result.best_fit,result.errorbars, label='Données ajustées')\n", "plt.xlabel(\"Dates (années)\")\n", "plt.ylabel(\"Concentration en CO$_{2}$ (ppm)\")\n", "plt.legend()\n", "plt.grid()\n", "plt.title('Ajustement des données par un modèle');" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "On peut alors constater que l'ajustement n'est pas si mauvais, la tendance semble bien être représentée, de même que les oscillations. Le modèle semble être suffisant pour pouvoir extrapoler la courbe dans le futur." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "2. Caractérisation de la tendance\n", "\n", "Nous avons dis précédemment que la tendance de la courbe était représentée ici par une fonction cubique à quelques constantes prêt. Par les résultats de l'ajustement nous trouvons comme expression de la tendance :\n", "\n", "$Tendance = 289.43+3.88.10^{-5}*(x-1870.92)^{2.99}$\n", "\n", "Nous pouvons superposer cette fonction à la courbe pour vérifier si l'expression correspond bien (résultat ci-dessous), et cela semble être le cas." ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "hideCode": true, "scrolled": false }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "tendance=FonctionModel(reduced_dates,result.best_values['t0'],0,0,0,0,result.best_values['E'],result.best_values['F'],result.best_values['G'],np.zeros(len(reduced_dates)))\n", "plt.plot(reduced_dates,reduced_data,label='Données réduites')\n", "plt.plot(reduced_dates,tendance, label='Tendance', linewidth=3)\n", "plt.xlabel(\"Dates (années)\")\n", "plt.ylabel(\"Concentration en CO$_{2}$ (ppm)\")\n", "plt.legend()\n", "plt.grid()\n", "plt.title('Représentation de la courbe de tendance modelisée');" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "3. Carctérisation des oscillations\n", "\n", "Concernant les oscillations, les ajustements de la courbe nous permettent d'exprimer le modèle comme :\n", "\n", "$Oscillation = 2.84*\\frac{cos(2\\pi*1*x)}{1!}+1.00*\\frac{cos(2\\pi*3*1*x)}{3!}+0.64*\\frac{cos(2\\pi*5*1*x)}{5!}+3.64.10^{-6}*\\frac{cos(2\\pi*7*1*x)}{7!}$\n", "\n", "On peut alors faire plusieurs remarques :\n", "- La fréquence est proche de 1/an.\n", "- La première harmonique est la plus prépondérante.\n", "- La dernière harmonique n'est pas indispensable pour cette modélisation puisque la constante est quasiment nulle.\n", "\n", "Traçons ces oscillations pour vérifier leur allure et leur pretinence (résultat ci-dessous). On remarque sur le graphique que les oscillations sont globalement de type triangle, ce qui est attendu, de même on remarque que la période est d'environ 1 an, ce qui correspond donc bien à notre fréquence ajustée ($fréquence=\\frac{1}{periode}$)." ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "hideCode": false }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "oscillation=FonctionModel(reduced_dates,result.best_values['t0'],result.best_values['A'],result.best_values['B'],result.best_values['C'],result.best_values['D'],np.zeros(len(reduced_dates)),0,0,np.ones(len(reduced_dates))*result.best_values['a'])\n", "plt.plot(reduced_dates[:100],oscillation[:100], label='Oscillations')\n", "plt.xlabel(\"Dates (années)\")\n", "plt.ylabel(\"Concentration en CO$_{2}$ (ppm)\")\n", "plt.legend()\n", "plt.grid()\n", "plt.title('Représentation des oscillations modelisées (Zoom)');" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "4. Estimation en 2025\n", "\n", "Maintenant que nous avons modélisé une fonction permettant de décrire nos données, nous pouvons aisément extrapoler la courbe aux dates futures. Par exemple, déduisons ici la concentration en CO$_{2}$ pour le 1$^{er}$ janvier 2025 (calcul via le modèle ci-dessous). Nous trouvons une concentration de 425.40 ppm. Bien évidemment, cette valeur est à relativiser puisque nous n'avons pas pris en compte les incertitudes liées à l'ajustement des variables du modèle, bien qu'elles semblent correctes. La valeur calculée est de toute façon du même ordre de grandeur que les données et supérieure à la dernière mesure de 2020, donc on peut quand même supposer notre intertitude sur le résultat faible." ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "423.6680255639254" ] }, "execution_count": 12, "metadata": {}, "output_type": "execute_result" } ], "source": [ "Estimation2025=FonctionModel(2025,result.best_values['t0'],result.best_values['A'],result.best_values['B'],result.best_values['C'],result.best_values['D'],result.best_values['E'],result.best_values['F'],result.best_values['G'],result.best_values['a'])\n", "Estimation2025\n" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "hide_code_all_hidden": false, "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.6.4" } }, "nbformat": 4, "nbformat_minor": 2 }