Data conserved in the coconut Data management Software as of 2019

By R. Bourdeix, 2019

CDM (Coconut Data Management) is a software intended to manage experimental data observed on collections or experimental fields of coconut over a long period following a regular schedule. It is based on the STANTECH manual which describes the organization of field planting, data gathering, and data analysis, along with the characters to be observed. Data management is made for every palm planted in the fields. Version 3 is able to manage the database containing these palms with their identification characters, along with data concerning immature vegetative observations, leaf morphology, stem measurements, flowering dates, inflorescence morphology, fruit and bunch return, fruit component analyses, and state of the palms. As the location of every palm in the field is recorded in the database, it is possible to draw maps of the fields with geographical representation of characters. Observations are dated with year and month and it is possible to record one value per character and per month. It is possible to execute powerful queries on the database, to export data into external files, and to make statistical analysis of widely used designs. The manual can be dowloaded by clicking here.
Figure 1. Illustration of the MAP fonction of the CDM software
From 2012, Dr R. Bourdeix have conserved a set of the PNG data that were the data that has been entered into the software CDM, Coconut Data Management. This set remains available on demand. The best (cleaned) files were also copied on a computer at SRS in the directory d;\backup Files\CDMV3\SRS.

To extract any data from CDM to excel, click on the dbf file name (for instance Fruit.dbf), it will open the Foxpro software;  then open the command window (fenetre then commande) and write:  “copy to temp.xls type xl5”; “copy(blank)to(blank)temp.xls(blank)type xl5”; it will generate an excel file named temp.xls with all the data

In 2012, the CDM software is used since 9 years (2003), Dr Faure used it first, then Wendy used it from 2004 without training, she used the CDM book. Before the data were keyboarded on excel files (we asked for an example of excel file, but did not get it). We analyzed the PNG data in CDM the 21/11/2012 using the Foxpro Software, which was used for writing both CDM and CGRD softwares.

Analysis of the file Bunch.dbf

In the file Bunch.dbf, 201853 records. Data start in 2003 up to 2011. 35000 to 45000 data per year from 2003 to 2006, then it reduced to about 15000 data per year in 2007 and 2008, then reduce again to less than one thousand per year. There are about 185 data without year of observation, and this is abnormal. It was empty data and we deleted them. We found about 122 data with more than 50 fruits harvested in one time that needs to be verified.  The instructions  “set filter to nb_bunch=0 and nb_fr<>0” on the Bunch.DBF give 5 records that needed to be corrected, so we replaced the number of bunches by 1.

Analysis of the file trees.dbf

The analysis of the file trees.dbf shows that observations were conducted on 9230 palms belonging to only 6 experiments, 702, 703, 704, 05, 706 (hybrid tests) and 709 (genebank). For 3412 records, the father was mentioned but not the mother, and we completed information regarding the mother.

Analysis of the file Fruit.dbf

The analysis of the file Fruit.dbf shows that 9204 fruits analysis were conducted. Fruit analysis conducted only in 4 fields, 061, 064, 071,  092 which are hybrid experiments. Fruit analysis conducted only from 2003 to 2009, with low number in 2007. Check 092 06 03 year 2003 month 3: 3 fruits sampled. Check 092 06 23 year 2003 month 10: 1 fruits sampled. We searched all the data where the weight of the whole fruit with husk was inferior to the weight  of dehusked fruit (coconut) and we found 297 (set filter to fr_totw<nut_totw). We searched all the data where the weight of the coconuts was inferior to the weight  of splitted coconut and we found 320 (set filter to nut_totw<spnut_totw). These errors discovered in 2012 were indicated as needing to be corrected.

Availalibility of the data on paper sheets

Normally, even if the data is computerized, the original paper datasheet were the data was registered should be preciously conserved. Where are presently these datasheet?

Putative errors detected in the CDM Software at SRS research Station

Please click here to see the lists established by Dr R. bourdeix in November 2012.