I am building ML models and have plotted the recall values at different % of complete data, like so:
The y axis represents the recall values, and the x axis shows the percentage of data completeness (so 0.6 complete data means that records with >40% missing data have been removed, 0.7 means that records with >30% missing data have been removed, etc.).
This is the code I used to generate this plot:
fig = plt.figure()
fig.suptitle("True Positive Rate")
ax = fig.add_subplot(111)
subsets=[0.5, 0.6, 0.7, 0.8, 0.9, 1]
ax.plot(subsets, recall_results, marker = "o", linestyle = "--")
ax.set_ylabel("True Positive Rate")
ax.set_xlabel("% complete data in samples")
plt.show()
To get an idea of how the model performance is changing when dropping records with different percentages of missing values in comparison with the original data, I want to add the baseline recall value at x point 0, so (after adding this value to my recall_values list) I changed my code to:
fig = plt.figure()
fig.suptitle("True Positive Rate")
ax = fig.add_subplot(111)
subsets=[0, 0.5, 0.6, 0.7, 0.8, 0.9, 1]
ax.plot(subsets, recall_results, marker = "o", linestyle = "--")
ax.set_ylabel("True Positive Rate")
ax.set_xlabel("% complete data in samples")
plt.show()
As you can see form the plot, the new recall value at point 0 has been added, but the x values have been changed from 0.5, 0.6, 0.7, 0.8, 0.9, 1 to 0.0, 0.2, 0.4, 0.6, 0.8, 1.0. I understand that the x points written in the new graph have equal distances from each other as opposed to jumping from 0 to 0.5.
There is nothing wrong with the new graph from a technical point of view, but I would rather the x values written be those that I have y (recall) values for [0, 0.5, 0.6, 0.7, 0.8, 0.9, 1]. Can anyone help me out?
Thanks!
ax.set_xticks(subsets)to only have ticks at those positions. Orax.xaxis.set_major_locator(matplotlib.ticker.MultipleLocator(0.1))to have ticks every0.1. Or changesubsetsto strings to have equally spaced ticks with only those numbers. – JohanC