ENVI_ Idl: create HDF5 file and write data (take writing GeoTIFF file to HDF file as an example) + detailed parsing

 

Catalog

 

1. Learning content

2. Knowledge reserve

3. Programming

---

1. Learning content

Article title ： Is to create by yourself HDF File and write data to it

——————————————————————————————————————————————————————————————————————————————————————

establish HDF5 Basic process of documents ：

1. Specify the path to create HDF5 file

2. Create groups that store the same type of data

3. Get the data type of the dataset id

4. Get the data space of the dataset id

5. Create a dataset id( Creating attributes is a similar step )

6. Write data to dataset

7. Close file

——————————————————————————————————————————————————————————————————————————————————————

2. Knowledge reserve

Here are some functions you may use .

——————————————————————————————————————————————————————————————————————————————————————

 

 —————————————————————————————————————————————————————————————————————————————————————

3. Programming

pro week_seven_study1
  ;  This program is used to solve the problem of creating hdf5 File and write data 
  ;  The data written here is some tiff File data 
  
  ;  The time when the program starts 
  start_time = systime(1)
  
  ;  route 
  in_dir = 'D:/IDL_program/experiment_data/chapter_4'
  out_dir = 'D:/IDL_program/experiment_data/chapter_4/hdf/'
  if file_test(out_dir) eq 0.0 then file_mkdir, out_dir  ;  If the directory does not exist, create 
  
  ;  obtain in_dir Under the air_quality dependent tiff File path and get tiff Number of files (count= return )
  tiff_path_list = file_search(in_dir, 'air_quality*.tiff', count=tiff_count)  ;  Because there are some tiff The document is not what we want , So you can't just *.tiff
  ;  obtain tiff The name of the document 
  tiff_name_list = file_basename(tiff_path_list, '.tiff')
  
  ;  establish hdf5 file ——》 adopt h5f_create( route ) Function creation , Return to the created hdf5 Of documents id
  create_path = out_dir + 'air_quality.he5'  ; hdf5 The suffix of the file is .he5
  h5_id = h5f_create(create_path)
  
  ;  establish HDF5 Group , For storing datasets ( upper tiff Body data of the file )——》 adopt h5g_create(Loc_id, Name) function (Loc_id It could be a file id It can also be a group id)
  ;  Which one do you want to put the group under , Which one do you import id, Here I put it directly in hdf5 Below file , So the introduction hdf5 Of documents id(h5_id),  The second parameter is the name of the Group 
  ds_group_id = h5g_create(h5_id, 'DataSets')
  
  ;  Here, a global attribute is created to describe this hdf5 A brief overview of the document 
  decribe = 'Here are some data about air quality'
  ;  Although you know that the data type of writing global attributes is string , But you need to let IDL Know and know , So you need to know the data type id
  ;  Incoming data returns data of type id——》h5t_idl_create(Data) Function passes in data , Return the data type of the data 
  global_att_type_id = h5t_idl_create(decribe)
  ;  In addition, all your incoming data is similar to excel In the form of , So you need to know how many rows and columns your data needs ——. That is, you need to tell IDL How much space is there to write data 
  ;  The size of the space that needs to be written is passed in, and so many data spaces are returned id——》h5s_create_simple( Incoming row and column ) Function returns... Of the data space id
  global_att_space_id = h5s_create_simple([1])  ;  As our string theory, one row and one column is enough , So you can directly [1], If necessary in the future x That's ok y Column then write [y, x]
  ;  establish hdf5 Properties of , Returns the id——》 adopt h5a_create(Loc_id, Datatype_id, Datasapce_id) function 
  ;  Where do you need to create the attribute , Then which position id, Here we create a global attribute , Obviously, this is a hdf5 Under the document , Therefore, the id, If you need to place attributes under a group , Then you should pass in the Group id
  global_att_id = h5a_create(h5_id, 'Decribe', global_att_type_id, global_att_space_id)
  ;  Now you can write data under a certain attribute , Because you have created this attribute and obtained its id——》 adopt h5a_write Function writes data to the attribute 
  h5a_write, global_att_id, decribe  ;  The first parameter is passed into the attribute that needs to write data id, The second parameter passes in the data to be written 
  
  
  
  ;  Now enter the loop and put each tiff The data of the file is extracted and written to the newly created hdf5 In the document 
  for tiff_i = 0, tiff_count - 1 do begin
    ;  Get the tiff File data , Get the subject information and geographic reference information 
    tiff_data = read_tiff(tiff_path_list[tiff_i], geotiff=tiff_geo)
    ;  To get the data size( Number of rows and columns )
    tiff_size = size(tiff_data)
    ;  Get the number of rows and columns of the data 
    tiff_column = tiff_size[1]
    tiff_row = tiff_size[2]
    ;  Get the resolution of the data 
    tiff_resolution = tiff_geo.(0)
    ;  Get the corner information of the data 
    tiff_geo = tiff_geo.(1)
    
    ;  Created earlier DataSets Create data sets in groups , Used to store the tiff data 
    ;  Similarly , You also need to get data types id, Get data space id, Create a dataset and get the id
    ;  Get the data type of the dataset id
    ds_type_id = h5t_idl_create(tiff_data)  ;  Incoming data about to be transferred into this dataset , Get the id
    ;  Get the data space of the dataset id
    ds_space_id = h5s_create_simple([tiff_column, tiff_row])
    ;  Create the dataset and get the id
    ds_id = h5d_create(ds_group_id, tiff_name_list[tiff_i], ds_type_id, ds_space_id, gzip=9)  ;  Here is an additional keyword parameter gzip, This indicates the compression degree of the incoming data , The greater the numerical , The higher the degree of compression , obtain hdf5 The file size will be correspondingly smaller , But the code will run longer 
    ;  Because the created dataset needs to be placed under the Group , So you need to pass in the Group id It's not a document id
    ;  Write data to the dataset 
    h5d_write, ds_id, tiff_data
    
    ;  Create dataset attributes in the dataset created above ( Here create a resolution attribute )
    ;  Get data type id
    res_type_id = h5t_idl_create(tiff_resolution)
    ;  Get data space id
    res_space_id = h5s_create_simple([3])
    ;  Create a resolution attribute and return the id
    res_id = h5a_create(ds_id, 'resolution', res_type_id, res_space_id)
    ;  Write data to this attribute 
    h5a_write, res_id, tiff_resolution
    
    ;  Allied , You can also create other attributes in the dataset to write corner information , There's no more demo
  endfor
  
  ;  Close the open hdf5 file 
  h5f_close, h5_id
  
  ;  The time when the program ends 
  stop_time = systime(1)
  ;  Give tips 
  print, ' Program time >>> ' + strcompress(string(stop_time - start_time)) + 's'
end

——————————————————————————————————————————————————————————————————————————————————————

I am fried eggplant , You're welcome