How to reduce a custom bin histogram for a region in Google Earth Engine?

Question

I'd like to reduce all the conties' data histogram with a custom bin, but what I got is a auto bin histogram, it seems ee.Reducer.fixedHistogram() can help, but I don't know how to pass the parameters(min,max,steps) for all conties.

// load precip data (mm, daily total): 365 images per year

var precipCollection = ee.ImageCollection('IDAHO_EPSCOR/GRIDMET')
                    .select('pr')   // select  precip band only
                    .filterDate('2017-01-01', '2017-12-31');
print(precipCollection, 'precipCollection');

// load precip data (mm, daily total): 365 images per year

var precipCollection = ee.ImageCollection('IDAHO_EPSCOR/GRIDMET')
                    .select('pr')   // select  precip band only
                    .filterDate('2017-01-01', '2017-12-31');
print(precipCollection);  

// reduce the image collection to one image by summing the 365 daily rasters

var annualPrecip = precipCollection.reduce(ee.Reducer.sum());
print(annualPrecip);

// equivalent shortcut syntax

var annualPrecip2 = precipCollection.sum();

// visualize annual precipitation
var precipPal = ['white','blue'] // store palette as variable               
Map.addLayer(annualPrecip, {min: 60, max: 3000, palette: precipPal}, 'precip');

// load regions: counties from a public fusion table, removing non-conus states // by using a custom filter

var nonCONUS = [2,15,60,66,69,72,78] // state FIPS codes that we don't want
var counties = ee.FeatureCollection('ft:1ZMnPbFshUI3qbk9XE0H7t1N5CjsEGyl8lZfWfVn4')
        .filter(ee.Filter.inList('STATEFP',nonCONUS).not());
print(counties, 'counties');

// visualize

Map.addLayer(counties,{},'counties');  
var steps = 128;

// get histogram precipitation values by county polygon(how to obtain custom bin histogram,such as 64/128/256 bins)

var countyPrecip = annualPrecip.reduceRegions({
  collection: counties,
  reducer: ee.Reducer.autoHistogram(),
  scale: 4000 // the resolution of the GRIDMET dataset
});


print(countyPrecip);

// drop .geo column (not needed if goal is tabular data)

var polyOut = countyPrecip.select(['.*'],null,false);

// add a new column for year to each feature in the feature collection

polyOut = polyOut.map(function(feature){
return feature.set('Year',2017);
});

// Table to Drive Export Example

Export.table.toDrive({
  collection: polyOut,
  description: 'GRIDMET_annual_precip_by_county',
  folder: 'GEE_geohackweek',
  fileFormat: 'CSV'
});  

Answer 1

You can specify the maximum number of bins for an autohistogram with the maxBuckets argument. Unless you have fewer values than that, you should get that many buckets.