<div dir="ltr"><div>Hello Moritz,</div><div><br></div>No worries. Thankful these modules are made available for newbies in RS like me and also happy these interactions are possible for learning. <div>Hope to get back soon after some adjustments. </div><div><br></div><div>Best,</div><div>Jamille</div></div><div class="gmail_extra"><br><div class="gmail_quote">On Thu, Jun 7, 2018 at 10:44 AM, Moritz Lennert <span dir="ltr"><<a href="mailto:mlennert@club.worldonline.be" target="_blank">mlennert@club.worldonline.be</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Thanks<span class=""><br>
<br>
On 07/06/18 15:17, Jamille Haarloo wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
The first 20+ lines of Stats_Training_Ben_test:<br>
<br>
cat,area,perimeter,compact_cir<wbr>cle,compact_square,fd,WV_Benat<wbr>imofo_1_min,WV_Benatimofo_1_<wbr>max,WV_Benatimofo_1_range,WV_B<wbr>enatimofo_1_mean,WV_Benatimofo<wbr>_1_stddev,WV_Benatimofo_1_<wbr>variance,WV_Benatimofo_1_<wbr>coeff_var,WV_Benatimofo_1_sum,<wbr>WV_Benatimofo_1_first_quart,<wbr>WV_Benatimofo_1_median,WV_Bena<wbr>timofo_1_third_quart,WV_Benati<wbr>mofo_2_min,WV_Benatimofo_2_<wbr>max,WV_Benatimofo_2_range,WV_B<wbr>enatimofo_2_mean,WV_Benatimofo<wbr>_2_stddev,WV_Benatimofo_2_<wbr>variance,WV_Benatimofo_2_<wbr>coeff_var,WV_Benatimofo_2_sum,<wbr>WV_Benatimofo_2_first_quart,<wbr>WV_Benatimofo_2_median,WV_Bena<wbr>timofo_2_third_quart,WV_Benati<wbr>mofo_3_min,WV_Benatimofo_3_<wbr>max,WV_Benatimofo_3_range,WV_B<wbr>enatimofo_3_mean,WV_Benatimofo<wbr>_3_stddev,WV_Benatimofo_3_<wbr>variance,WV_Benatimofo_3_<wbr>coeff_var,WV_Benatimofo_3_sum,<wbr>WV_Benatimofo_3_first_quart,<wbr>WV_Benatimofo_3_median,WV_Bena<wbr>timofo_3_third_quart,WV_Benati<wbr>mofo_4_min,WV_Benatimofo_4_<wbr>max,WV_Benatimofo_4_range,WV_B<wbr>enatimofo_4_mean,WV_Benatimofo<wbr>_4_stddev,WV_Benatimofo_4_<wbr>variance,WV_Benatimofo_4_<wbr>coeff_var,WV_Benatimofo_4_sum,<wbr>WV_Benatimofo_4_first_quart,<wbr>WV_Benatimofo_4_median,WV_Bena<wbr>timofo_4_third_quart<br>
1144,3832.000000,1256.000000,5<wbr>.723635,0.197144,1.729624,13,7<wbr>6,63,46.4097077244259,9.984549<wbr>11351384,99.69122100017,21.<wbr>5139237092391,177842,40,47,53,<wbr>40,138,98,90.2687891440501,15.<wbr>2500825418009,232.565017531741<wbr>,16.8940812061464,345910,81,<wbr>92,100,15,61,46,40.85829853862<wbr>21,7.82663897784868,61.2562776<wbr>895802,19.1555675536767,<wbr>156569,36,42,47,28,124,96,68.<wbr>4253653444676,13.5774536655369<wbr>,184.347248039801,19.<wbr>8427200164517,262206,59,68,77<br>
1145,12092.000000,2282.000000,<wbr>5.854120,0.192750,1.645226,13,<wbr>94,81,51.386288455177,10.52943<wbr>76761475,110.869057775874,20.<wbr>4907534532914,621363,45,52,59,<wbr>21,220,199,114.230731061859,<wbr>23.3590328249442,545.644414516<wbr>822,20.4489917973953,1381278,<wbr>101,114,128,7,76,69,46.<wbr>4219318557724,8.42747122371732<wbr>,71.0222712265835,18.154072626<wbr>4915,561334,42,48,52,17,198,<wbr>181,97.2732385047966,22.<wbr>492313569247,505.904169697333,<wbr>23.1228176577445,1176228,84,<wbr>97,110<br>
<br>
</blockquote></span>
[...]<span class=""><br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
---------------------<br>
All the lines of the output of v.db.select Training_Ben2@haarlooj_Ben_Tes<wbr>t:<br>
<br>
cat|id|Type|code<br>
1|4|B29|18<br>
2|5|B31|19<br>
3|3|B28|17<br>
</blockquote>
<br>
<br></span>
Again a lack of clear documentation on my side: both the training and the segment info should contain the same attributes, with only additional one column ('code' in your case) present in the training data.<br>
<br>
It should be possible to do this differently, i.e. provide the module with the features of all segments, and only the id/cat of each training segment with the relevant class and have the module merge the two, but this is not implemented, yet.<br>
<br>
I also just notice that you have the word 'Training' in both names.<br>
<br>
The segment_file/segment_map contains the info (cat + all feature variables) of all segments you wish to classify, either in the form of a csv file or in the form of a vector map with the info in the attribute table.<br>
<br>
The training_file/training_map contains the info (cat + all feature variables + class) of the training data. Often this is an extract of the former, but not necessarily.<br>
<br>
All columns in the training file have to be present in the segment file, except for the class column (your 'code').<br>
<br>
Sorry for the lack of docs. This module has mostly been used internally here and so we are not always aware of the unclear and missing parts. Having your feedback has been very useful !<span class="HOEnZb"><font color="#888888"><br>
<br>
Moritz<br>
<br>
</font></span></blockquote></div><br></div>