CDO

This is my last line of my code :

data = xr.open_dataset("./mrros_1850-2014_ymean.nc")
mrros = data.mrros.data

dataa = xr.open_dataset("./tos_1850-2014_ymean.nc")
tos = dataa.tos.data

plt.figure(figsize=(24, 12))

plt.title('Ocean surface temperature in function of the surface runoff for September to November from 1850 to 2014', size=20)
plt.xlabel('Surface runoff (kg.m-2.s-1)')
plt.ylabel('Ocean surface temperature($^\degree$C)')
plt.scatter(mrros, tos)

Maybe I should create a new dataset with regress data and add this to my plot ?

I not sure of what is the best and the more correct way to do this.

I would recommend to use Numpy's polyfit to generate the trend line for the scatter plot.

Short example with random numbers

import numpy as np
import matplotlib.pyplot as plt

x = np.random.uniform(1, 200,100)
y = np.random.uniform(50, 100,100)

m, b = np.polyfit(x, y, 1)

fig = plt.figure(figsize=(12,6))
ax = fig.add_subplot(111)

ax.scatter(x, y)
ax.plot(x, m*x + b)

If I want to use my data, can I only do :

tos = np.array(tos)
mrros = np.array(mrros)

and after replace x and y by mrros and tos ?

I tried and I have :

TypeError: expected 1D vector for x

This is a Python error and not CDO, besides without the data files you use it is not really possible to help.

Does your short script run without the regression part?

Yes my script run before the regression part.

This my data, I plot : tos = f(mrros), I tried to add the regression part but it doesn't work, maybe because of the size of my data (it is netCDF file).

I just want to add a linear regression line do this plot. I also tried with the tool 'linear regression" but impossible to install this package.

I understand that this does not concern cdo anymore so it has no place on this forum. Anyway, I thank you for your precious help so far !

Well, as Dennis Shea likes to say: 'Have a look at your data!'.

The data.mmros and dataa.tos are multi-dimensional and you have to get rid of the one-size dimensions with the squeeze() method.

import matplotlib.pyplot as plt
import numpy as np
import xarray as xr

data = xr.open_dataset("/Users/k204045/Downloads/mrros_1850-2014_ymean.nc")
mrros = data.mrros.squeeze()

dataa = xr.open_dataset("/Users/k204045/Downloads/tos_1850-2014_ymean.nc")
tos = dataa.tos.squeeze()

#-- compute slope and intercept
m, b = np.polyfit(mrros, tos, 1)

#-- create the plot
fig, ax = plt.subplots(figsize=(12,6))

plt.title('Ocean surface temperature in function of the surface runoff for September to November from 1850 to 2014', size=12)

plt.xlabel('Surface runoff (kg.m-2.s-1)')
plt.ylabel('Ocean surface temperature($^\degree$C)')

plt.scatter(mrros, tos)

ax.plot(mrros, m*mrros + b)

​plt.savefig('plot_scatter_plus_trend.png', bbox_inches='tight', facecolor='white', dpi=200)