Re: [ferret_users] How to map areas frequently characterized with low values

[milik a <milikagrow@xxxxxxxxx>]

This is great, thank you. I could get the desired map in percentages. Thank you Ryo.

regards

Milika

On Mon, Nov 19, 2018 at 8:04 AM Ryo Furue <furue@xxxxxxxxxx> wrote:

Dear Milik

On Sat, Nov 17, 2018 at 12:44 AM milik a <milikagrow@xxxxxxxxx> wrote:

I have a 5-year daily data of chlorophyll for the globe. I want to map the areas which are frequently characterised by  low values throughout year and areas characterised by infrequent low values. For this, first I did frequency_histogram where  I can find what are the predominant high/low values in the data.

But i want a fill map spatially showing the annual percentage of areas frequently occurring low values or high values.

I think you can combine the following three ideas to achieve what you need.

1) Mask.  For example, this mask

let mask = if yourvar lt threshold then 1

is 1 when yourvar < threshold and is undefined otherwise.

2) "@ transformation" to count the number of "good" values or undefined values.  Look into the official Ferret manual to find out which transformation you need.

3) Get the total size of you data points.  For example, 

let size = `yourvar,return=lend`

gives you the total size of the time axis.  So, the percentage you want is

(the number of good points) / size * 100

Hope this helps,

Ryo