"# Sujet 1: Concentration de CO2 dans l'atmosphère depuis 1952"
"# Downloading the data and first processing"
]
},
{
...
...
@@ -1943,6 +1943,7 @@
"\"CO2 seasonally adjust filled\" and \"CO2 seasonally filled\" are identical to Columns \"CO2 seasonally\" and \"CO2 seasonally adjust\" except that the missing values from have been filled with values from Columns \"fit\" and \"seasonally adjusted fit\"\n",
"\n",
"Moreover, it is said that the missing data are completed with the value -99.99\n",
"# Processing for a better representation\n",
"\n",
"The month and year can be then converted to panda Period.\n",
"\n",
...
...
@@ -1969,6 +1970,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"# Seasonal and interannual variations\n",
"In order to highlight the seasonal and interannual variations, only the raw concentrations stored in 'CO2 seasonally\" and \"CO2 seasonally adjust\" are enough. The first one is the raw concentrations and contains both the seasonal and interannual variations. The second one is a adjustement made in order to get rid of the seasonal variations."
]
},
...
...
@@ -2056,7 +2058,7 @@
"metadata": {},
"source": [
"This seasonal variations are rather close to a cosinus function with a one year period. In the other hand, the interannual variations are nearly linear. \n",
"\n",
"# Linear model for the interannual variations\n",
"Hence, a linear regression is going to be performed in order to get the slope and the intercept"
]
},
...
...
@@ -2113,6 +2115,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"# Quadratic model for the interannual variations \n",
"It actually seems that the interannual variations are more like a second order polynomial function"
