Issue on page /user-guide/time-series.html

[gagebeni]

Hi all
I want to know if it is possible to index on the lat two points having different coordinates like this
mask_lon = (data_sst.lon> = x1) & (data_sst.lon <= x0)
mask_lat = (data_sst.lat == y0, y1) & (data_sst.lat == y0, y1)
and then extract the transect resulting from this indexing like this:
transect = data_sst.SST.where (mask_lon & mask_lat, drop = True)
Thank you in advance.

[max-sixty]

Could you please review the issue template re a MVCE?

[gagebeni]

i am not yet experimenting enough on python. My problem is that I wrote some code to draw a line using a colon at the beginning and at the end of my line I used numpy exclusively and since it is an xrarray array it takes me a long time. My data is all of dimensions lat, lon, SST, (1501,2001), and here is my code:

    import xarray as xr
    import glob
    from numpy import *
    import numpy as np
    import matplotlib.pyplot as plt
    from scipy.signal import get_window
    import math
    import pandas as pd
    import numpy as np
    
    filepaths_TEMP = "./ 2014 / MUR-SST_2014__Day_027_27-Jan-2014.nc"
    
    data_sst = xr.open_dataset (filepaths_TEMP)
    lon_sst = data_sst.variables ['lon'] [:]
    lat_sst = data_sst.variables ['lat'] [:] #
    sst = data_sst.variables ['SST'] [:]
    #% to draw the line
    latlim = [50, 50.5];
    lonlim = [-10.2];
    n1 = 1200; #% number of point along profile
    delta_lon = -10 - (2);
    delta_lat = 50 - 50.5;
    
    #cd (input_path1)
    #matfiles = dir (fullfile ('*. nc'));
    # %%
    Flat = latlim [1] + np.arange (0, n1) * delta_lat / n1
    Plon = lonlim [1] + np.arange (0, n1) * delta_lon / n1
    transect = np.zeros ((1200,1))
    for ii in range (0, len (Flat)):
          
        lat_S1 = Flat [ii];
        lon_S1 = Plon [ii];
        a_1 = abs (lat_S1-lat_sst);
        a_2 = abs (lon_S1-lon_sst);
        a_3 = (a_1 + a_2);
        m_3 = a_3 [:]. min ()
        [px, py] = np.where (m_3 == a_3);
        px = px [0];
        py = py [0]; # clear a_1 a_2 a_3 m_3
        transect [ii] = sst [px, py];

My question is, isn't there a way to make it simpler and faster, especially thank you very much in advance.

[dcherian]

I recommend looking at .query and .isin.

transect = data_sst.SST.where (mask_lon & mask_lat, drop = True) looks like it could work. Did you try it?

This webpage might also be useful: https://xarray.pydata.org/en/stable/user-guide/indexing.html#more-advanced-indexing