Matplotlib segmented Plot-CodePudding

I have the following dataset:

   df = pd.DataFrame ({"a": [1,2,3,4,5,6,7,8,9,1,11,12,13,14,15,16,17,18,19,20], 
                'b':[1,2,3,4,50,60,70,8,9,10,110,120,130,140,150,16,17,18,19,20],
                'c':[np.nan,2.2,3.4,np.nan,40.9,60.2,np.nan,8.2,8.9,10.1,np.nan,120.2,
                     130.07,140.23,np.nan,16.054,17.20,18.1,np.nan,20.1],
               'd': [100, np.nan,np.nan, 500,np.nan, np.nan,500, 
                     np.nan,np.nan,np.nan,100, np.nan,np.nan, np.nan,500,
                     np.nan,np.nan, np.nan,100,np.nan  ]}
              )

I am trying to plot the data based on the following conditions:

Between 100 to the next 100 in column 'd' I want to have one plot having column 'a' in the x axis, and scatterplot of column 'b' and line plot of 'c' in the y axis.
That is I will be having 3 different plots. First one from index 0 to 10, second one from index 10 to index 18, third one from 18 to 20. (I can generate this using for loop)
Within each plot I want segmented lineplot based on the location 500 value in column 'd',i.e., for the first plot from index 0-3 one lineplot, from index 3-6 another and from index 6-10 another lineplot.( I can't make the segmented lineplot)

I am using the following codes:

    index = index   [len(df)]
    index1 = index1   [len(df)]

   for k in range (len(index)-1):
      x = df['a'][index[k]   1:index[k 1]]
      y = df['c'][index[k]  1:index[k 1]]
      y1 = df['b'][index[k]  1:index[k 1]]

      plt.scatter(x, y)
      plt.plot(x, y1)
      plt.savefig('plot'  str(k 1000)  '.png')
      plt.clf()

My first plot look like this: (But want to have three segmented lineplot not the continuous one (that is line from index 0-3 should not be connected with 3-6 and so on)

Sorry for the rather long question and thx:)

CodePudding user response：

The expected output is unclear, but here is a general strategy to split your dataset in groups with help of

option 3

ax = plt.subplot()
group = df['d'].eq(100).cumsum()
for name, g in df.groupby(group):
    g = g.reset_index()
    ax.scatter(g.index 1, g['c'])
    ax.plot(g.index 1, g['b'])