Investigate identifying new insertions using Median absolute deviation

PlantandFoodResearch / TEFingerprint

A Python3 library for transposon fingerprinting

MIT License

0 stars 2 forks source link

Investigate identifying new insertions using Median absolute deviation #124

Closed timothymillar closed 5 years ago

timothymillar commented 5 years ago

plot Median read-tip count by Median absolute deviation

example code

mad = np.median(np.abs(counts - np.expand_dims(np.median(counts, axis=1), 1)), axis=1)
median_count = np.median(counts, axis=1)
plt.scatter(median_count, mad)
plt.show()

mad_plot

timothymillar commented 5 years ago

MAD doen't work well - should check same plot with other metrics?