Using numpy for elevation statistics and visualization

Let's watch a video of the common evils ：

After watching the , Shrimp God thought , Isn't that what hadoop Basic skills inside wordcount Well …… But don't hadoop, Also need not Spark, use Numpy How to do it ？ Do what you say , Let's see the effect first ：

Although functions are often easy to implement , However, the effect of not being able to support others is flawed , So there is the following dialogue ：

Point directly to the topic in one sentence …… Millions later , Begging yard farmer …… Too real .

Let's briefly talk about the basic principle of the whole function realization ：

First , If you don't paint , It is the simplest elevation frequency statistics , If you like to do it with a dictionary , It's also easy , Like this ：

Here's divided by 10 After rounding, multiply by 10, It is used for reclassification , Give up all the ten places and all the places ： such as 4530, Round it down to 4500.

But such a calculation actually used 5.54s…… Isn't it a double cycle , So slowly , There is a way to hurry up ？ The answer, of course, is wooden , For example, direct use numpy Built in functions for ：

Oh my god ！84 millisecond …… use numpy The speed of , was Python Circularly calculated 65 times …… Don't write the statistical process by yourself , Sure enough Python Using bags is the king .

numpy Of unique function , The unique value statistics can be carried out directly , After the statistics are finished , You can draw the histogram directly ：

Actually so far , Even if the elevation statistics are finished ……

Color ！ Where is the color ？ Don't say color , Let's finish the work in a rough way ！！

ok , The following to continue …… Set the color of matplotlib A basic skill of , Directly through cmap Color mapping , You can render directly , For example, for terrain , We use different colors to map ：

gist_earth, The so-called earth color ：

Or directly use gray color ：

matplotlib Logo color jet（ Hot and cold color ）：

cmap The principle of is to directly generate a linear hierarchical color builder ： Well, I picked up the name myself , The official name is “LinearSegmentedColormap ”, The function is based on the color name you set , Generate a series of color values . So we can directly from this builder , To get the color we need ：

Now I'll start with jet Inside , obtain 10 One color ：

This is a rgb+ The value of transparency , Of course , We can also change the standard color number to hexadecimal ：

Of course , You can also use gradients ：

Know these methods , You can naturally get the rendering colors of different elevations , Then get the histogram , All the source codes are as follows ：

If you need source code , You can go to shrimp God's new code warehouse to find .

Code reference address , Send directly to official account 6, You can get , The above codes and data , All in PythonDemo/021 Elevation statistics and visualization .