Unable to generate a line trace together with a scatter trace in the same figure

đź“Š Plotly Python
1

Hi, I am having great difficulty in getting a regression line to be drawn on the same figure as a scatter trace. I am just really puzzled because there is nothing complicated here, and I am following quite closely with the line and scatter examples.

Here is the code that I am trying to use to generate the plot:

datapts = go.Scatter(x = data["Population"], y=data["Profit"], mode = "markers",
                hoveron = "points")
reg_line = go.Scatter(x = pop_range, y = profit_predicted, mode = "lines",
                     hoveron="points")

plotly.offline.iplot({
    "data": [datapts,reg_line],
    "layout": go.Layout(title="Relationship between Population and Profit", xaxis = dict(title="Population of City in 10,000s"), 
                     yaxis = dict(title="Profit in $10,000s"))
})

The figure generated just looks like:

newplot.png700Ă—450 19.8 KB

As you can see, the line, trace 1, just does not show up at all. To make it possible to help me, these are what the arrays look like:

data["Population"][0:5]  # 97 elements long
0    6.1101
1    5.5277
2    8.5186
3    7.0032
4    5.8598
Name: Population, dtype: float64

data["Profit"][0:5]  # 97 elements long
0    17.5920
1     9.1302
2    13.6620
3    11.8540
4     6.8233
Name: Profit, dtype: float64

pop_range[0:5]  # 100 elements long
array([ 5.0269    ,  5.20039596,  5.37389192,  5.54738788,  5.72088384])

profit_predicted[0:5]  # 100 elements long
matrix([[ 2.23289546],
        [ 2.43525461],
        [ 2.63761377],
        [ 2.83997293],
        [ 3.04233208]])

I’m at my wits end. I’ve tried casting the profit_predicted matrix to a 1 dimensional flattened numpy array, but the bug does not go away, so I really do not know what’s the cause here.

Thank you.

2

@kohaugustine It doesn’t plot anything because you are using np.matrix, and Plotly doesn’t recognize this type of data. Convert it to a flattened numpy array and the line(s) will show up.
Paste here, please, your code with converted profit_predicted to np.array to see why the line wasn’t plotted.

3

@empet the code that I used to flatten profit_predicted was: np.array(profit_predicted).T. The result from that still did not work when I passed it into the y parameter.

Your reply gave me further confidence that the root cause for the line not showing up has to do with the unsuitable array type, so I tried to find other more certain ways that I could flatten profit_predicted, and realized that there is an numpy.ndarray.flatten() method (pardon me, I’m still new to numpy) and found that right ways are np.asarray(profit_predicted).flatten() or np.array(profit_predicted).flatten(). Both of these ways give a resulting array of:

np.asarray(profit_predicted).flatten()[0:5]
array([ 2.23289546,  2.43525461,  2.63761377,  2.83997293,  3.04233208])

and when I looked back to my original code, I realized that there was a feature in the original output that I overlooked:

np.array(profit_predicted).T[0:5]
array([[  2.23289546,   2.43525461,   2.63761377,   2.83997293,
          3.04233208,   3.24469124,   3.44705039,   3.64940955,
          3.8517687 ,   4.05412786,   4.25648701,   4.45884617,
          4.66120532,   4.86356448,   5.06592363,   5.26828279,
          5.47064194,   5.6730011 ,   5.87536025,   6.07771941,
......
     17.61219126,  17.81455041,  18.01690957,  18.21926872,
     18.42162788,  18.62398703,  18.82634619,  19.02870534,
     19.2310645 ,  19.43342365,  19.63578281,  19.83814196,
     20.04050112,  20.24286027,  20.44521943,  20.64757858,
     20.84993774,  21.05229689,  21.25465605,  21.4570152 ,
     21.65937436,  21.86173351,  22.06409267,  22.26645183]])

now its totally clear to me why my “flattening” also did not work!

I’m now able to generate the line:

It was quite a silly mistake, so I really appreciate your help! Thank you very much!