2
$\begingroup$

The input file looks like this, and the complete file can be found here:

OG0000008 NbL07g11380.1 NbL19g07810.1 NbL19g09170.1 NbL19g19070.1 NbQ01g01670.1 NbQ01g03330.1 NbQ01g04070.1 NbQ01g04670.1 NbQ01g05120.1 NbQ01g05870.1 NbQ01g06940.1 NbQ01g07580.1 NbQ01g08860.1 NbQ01g10050.1 NbQ01g10360.1 NbQ01g14200.1 NbQ01g14790.1 NbQ01g16080.1 NbQ01g17760.1 NbQ01g19270.1 NbQ01g19310.1 NbQ01g19390.1 NbQ01g21260.1 NbQ01g21330.1 NbQ01g21740.1 NbQ01g21910.1 NbQ01g23100.1 NbQ01g24620.1 NbQ01g25340.1 NbQ01g26060.1 NbQ01g26320.1 NbQ02g00750.1 NbQ02g03100.1 NbQ02g03420.1 NbQ02g03610.1 NbQ02g03680.1 NbQ02g05120.1 NbQ02g07460.1 NbQ02g08170.1 NbQ02g08330.1 NbQ02g09220.1 NbQ02g09400.1 NbQ02g10620.1 NbQ02g11310.1 NbQ02g14330.1 NbQ02g14460.1 NbQ02g14520.1 NbQ02g15320.1 NbQ02g17090.1 NbQ02g17130.1 NbQ02g20290.1 NbQ02g23070.1 NbQ02g23420.1 NbQ02g24450.1 NbQ02g24480.1 NbQ02g26700.1 NbQ03g00830.1 NbQ03g01970.1 NbQ03g04460.1 NbQ03g06900.1 NbQ03g09530.1 NbQ03g10620.1 NbQ03g12760.1 NbQ03g13450.1 NbQ03g15540.1 NbQ03g15640.1 NbQ03g17180.1 NbQ03g20740.1 NbQ03g21510.1 NbQ03g24670.1 NbQ04g01350.1 NbQ04g01720.1 NbQ04g08420.1 NbQ04g09090.1 NbQ04g10450.1 NbQ04g11470.1 NbQ04g12120.1 NbQ04g14130.1 NbQ04g15440.1 NbQ04g15860.1 NbQ04g16450.1 NbQ04g16620.1 NbQ04g17760.1 NbQ04g19040.1 NbQ04g20020.1 NbQ05g03320.1 NbQ05g04660.1 NbQ05g05970.1 NbQ05g07500.1 NbQ05g08900.1 NbQ05g09760.1 NbQ05g10830.1 NbQ05g11150.1 NbQ05g11340.1 NbQ05g11510.1 NbQ05g11530.1 NbQ05g11780.1 NbQ05g16980.1 NbQ05g18190.1 NbQ05g21710.1 NbQ05g23400.1 NbQ06g01110.1 NbQ06g01430.1 NbQ06g04200.1 NbQ06g04440.1 NbQ06g05330.1 NbQ06g05770.1 NbQ06g05820.1 NbQ06g06700.1 NbQ06g08620.1 NbQ06g09190.1 NbQ06g10460.1 NbQ06g15220.1 NbQ06g15330.1 NbQ06g15700.1 NbQ06g16320.1 NbQ06g16590.1 NbQ06g17590.1 NbQ06g17670.1 NbQ06g20050.1 NbQ07g01030.1 NbQ07g02010.1 NbQ07g04350.1 NbQ07g04900.1 NbQ07g05610.1 NbQ07g06200.1 NbQ07g07110.1 NbQ07g07690.1 NbQ07g08640.1 NbQ07g10390.1 NbQ07g11920.1 NbQ07g14130.1 NbQ07g15590.1 NbQ07g15620.1 NbQ07g16910.1 NbQ07g17130.1 NbQ07g17950.1 NbQ08g00060.1 NbQ08g02240.1 NbQ08g02300.1 NbQ08g02310.1 NbQ08g03290.1 NbQ08g05330.1 NbQ08g09280.1 NbQ08g14890.1 NbQ08g15820.1 NbQ08g15950.1 NbQ08g19830.1 NbQ08g20150.1 NbQ08g22050.1 NbQ08g22620.1 NbQ09g02100.1 NbQ09g02620.1 NbQ09g03950.1 NbQ09g04200.1 NbQ09g06040.1 NbQ09g06640.1 NbQ09g08160.1 NbQ09g08330.1 NbQ09g09660.1 NbQ09g11220.1 NbQ09g13860.1 NbQ09g15180.1 NbQ09g15310.1 NbQ09g16530.1 NbQ09g17900.1 NbQ09g18100.1 NbQ09g18720.1 NbQ09g19280.1 NbQ09g21840.1 NbQ10g00480.1 NbQ10g01350.1 NbQ10g02870.1 NbQ10g03640.1 NbQ10g03730.1 NbQ10g08070.1 NbQ10g09510.1 NbQ10g11010.1 NbQ10g11760.1 NbQ10g12050.1 NbQ10g12060.1 NbQ10g12910.1 NbQ10g19200.1 NbQ10g19930.1 NbQ10g20390.1 NbQ10g20730.1 NbQ10g21080.1 NbQ10g21140.1 NbQ10g24010.1 NbQ11g00310.1 NbQ11g01210.1 NbQ11g01370.1 NbQ11g04610.1 NbQ11g04800.1 NbQ11g06060.1 NbQ11g07820.1 NbQ11g08390.1 NbQ11g09100.1 NbQ11g09350.1 NbQ11g13660.1 NbQ11g13930.1 NbQ11g16260.1 NbQ11g17360.1 NbQ11g18430.1 NbQ11g21080.1 NbQ11g23280.1 NbQ11g23990.1 NbQ11g25050.1 NbQ12g03770.1 NbQ12g04850.1 NbQ12g07340.1 NbQ12g09080.1 NbQ12g10820.1 NbQ12g12070.1 NbQ12g14750.1 NbQ12g15000.1 NbQ12g15230.1 NbQ12g20380.1 NbQ12g21080.1 NbQ12g21830.1 NbQ12g23960.1 NbQ13g01300.1 NbQ13g02350.1 NbQ13g03860.1 NbQ13g04410.1 NbQ13g08800.1 NbQ13g09850.1 NbQ13g10370.1 NbQ13g11700.1 NbQ13g12420.1 NbQ13g15780.1 NbQ13g16040.1 NbQ13g23160.1 NbQ13g24120.1 NbQ13g24540.1 NbQ13g25080.1 NbQ13g25490.1 NbQ13g28240.1 NbQ13g29770.1 NbQ14g01070.1 NbQ14g03950.1 NbQ14g05360.1 NbQ14g05410.1 NbQ14g06880.1 NbQ14g07270.1 NbQ14g07500.1 NbQ14g10290.1 NbQ14g10770.1 NbQ14g14320.1 NbQ14g17890.1 NbQ14g18710.1 NbQ14g20960.1 NbQ14g22890.1 NbQ15g00150.1 NbQ15g02300.1 NbQ15g02330.1 NbQ15g02350.1 NbQ15g03230.1 NbQ15g06190.1 NbQ15g07120.1 NbQ15g07750.1 NbQ15g09000.1 NbQ15g09050.1 NbQ15g11920.1 NbQ15g12650.1 NbQ15g12840.1 NbQ15g15670.1 NbQ15g15930.1 NbQ15g18670.1 NbQ15g19070.1 NbQ15g20620.1 NbQ15g22880.1 NbQ15g23000.1 NbQ15g26060.1 NbQ16g00880.1 NbQ16g04360.1 NbQ16g06490.1 NbQ16g09100.1 NbQ16g11020.1 NbQ16g11560.1 NbQ16g13810.1 NbQ16g13820.1 NbQ16g17040.1 NbQ16g17130.1 NbQ16g17340.1 NbQ16g18390.1 NbQ16g18430.1 NbQ16g23100.1 NbQ16g23570.1 NbQ16g24270.1 NbQ16g25200.1 NbQ16g25830.1 NbQ16g25880.1 NbQ16g25990.1 NbQ16g26610.1 NbQ16g26660.1 NbQ16g28010.1 NbQ16g28180.1 NbQ17g01150.1 NbQ17g01180.1 NbQ17g01570.1 NbQ17g01950.1 NbQ17g05460.1 NbQ17g05540.1 NbQ17g05980.1 NbQ17g07990.1 NbQ17g08300.1 NbQ17g09330.1 NbQ17g09400.1 NbQ17g10090.1 NbQ17g11220.1 NbQ17g13030.1 NbQ17g15460.1 NbQ17g16690.1 NbQ17g20980.1 NbQ17g22370.1 NbQ17g25040.1 NbQ17g28730.1 NbQ18g02140.1 NbQ18g02740.1 NbQ18g05440.1 NbQ18g06120.1 NbQ18g07470.1 NbQ18g12320.1 NbQ18g12530.1 NbQ18g12850.1 NbQ18g13840.1 NbQ18g14420.1 NbQ18g14930.1 NbQ18g15730.1 NbQ18g17750.1 NbQ18g17850.1 NbQ18g21060.1 NbQ19g01040.1 NbQ19g05480.1 NbQ19g06450.1 NbQ19g06510.1 NbQ19g08330.1 NbQ19g11840.1 NbQ19g11880.1 NbQ19g13750.1 NbQ19g14190.1 NbQ19g14210.1 NbQ19g14920.1 NbQ19g18540.1 NbQ19g19870.1 NbQ19g21020.1 NbQ19g21220.1 NbQ19g22080.1 NbQ19g22800.1 NbQ19g24690.1 NbQ19g24730.1 rna19561
OG0000001 Capann_59V1aChr01g048170.1 NbL01g00020.1 NbL01g00940.1 NbL01g02330.1 NbL01g03550.1 NbL01g03650.1 NbL01g04410.1 NbL01g04920.1 NbL01g06850.1 NbL01g16120.1 NbL01g19150.1 NbL01g20140.1 NbL01g20930.1 NbL01g22230.1 NbL01g24190.1 NbL01g24280.1 NbL01g24300.1 NbL02g00570.1 NbL02g00900.1 NbL02g01270.1 NbL02g02110.1 NbL02g02210.1 NbL02g02470.1 NbL02g03180.1 NbL02g04740.1 NbL02g04750.1 NbL02g06120.1 NbL02g06860.1 NbL02g07280.1 NbL02g07680.1 NbL02g07740.1 NbL02g09780.1 NbL02g11320.1 NbL02g12670.1 NbL02g13080.1 NbL02g14050.1 NbL02g14190.1 NbL02g15010.1 NbL02g15890.1 NbL02g16190.1 NbL02g16730.1 NbL02g17070.1 NbL02g17360.1 NbL02g18820.1 NbL02g19340.1 NbL02g20100.1 NbL02g23950.1 NbL02g24800.1 NbL03g01610.1 NbL03g01680.1 NbL03g01890.1 NbL03g02230.1 NbL03g02600.1 NbL03g03410.1 NbL03g04990.1 NbL03g05400.1 NbL03g08030.1 NbL03g08250.1 NbL03g08690.1 NbL03g10230.1 NbL03g11060.1 NbL03g13030.1 NbL03g14960.1 NbL03g15110.1 NbL03g16690.1 NbL03g16900.1 NbL03g18260.1 NbL03g18950.1 NbL03g21180.1 NbL03g21210.1 NbL03g21530.1 NbL03g22960.1 NbL03g24430.1 NbL04g01140.1 NbL04g01490.1 NbL04g02030.1 NbL04g02560.1 NbL04g03700.1 NbL04g04160.1 NbL04g05240.1 NbL04g05420.1 NbL04g05850.1 NbL04g12420.1 NbL04g12640.1 NbL04g13650.1 NbL04g13780.1 NbL04g14310.1 NbL04g16260.1 NbL04g17750.1 NbL04g18380.1 NbL04g18870.1 NbL04g19030.1 NbL04g19630.1 NbL05g00320.1 NbL05g03060.1 NbL05g03300.1 NbL05g04060.1 NbL05g07620.1 NbL05g08630.1 NbL05g09580.1 NbL05g10060.1 NbL05g11400.1 NbL05g12280.1 NbL05g13170.1 NbL05g16020.1 NbL05g17530.1 NbL05g17730.1 NbL05g18340.1 NbL05g18590.1 NbL05g18600.1 NbL05g19640.1 NbL05g20640.1 NbL05g21000.1 NbL05g21640.1 NbL05g22610.1 NbL06g00640.1 NbL06g00660.1 NbL06g02210.1 NbL06g03150.1 NbL06g03680.1 NbL06g04910.1 NbL06g07950.1 NbL06g09970.1 NbL06g11480.1 NbL06g12220.1 NbL06g12400.1 NbL06g12460.1 NbL06g12850.1 NbL06g13120.1 NbL06g14450.1 NbL06g14780.1 NbL06g16990.1 NbL06g17200.1 NbL06g17760.1 NbL06g20380.1 NbL07g02100.1 NbL07g02540.1 NbL07g02970.1 NbL07g03110.1 NbL07g04840.1 NbL07g05350.1 NbL07g06580.1 NbL07g07530.1 NbL07g08450.1 NbL07g09380.1 NbL07g09870.1 NbL07g10730.1 NbL07g10850.1 NbL07g11080.1 NbL07g12450.1 NbL07g12710.1 NbL07g13110.1 NbL07g13920.1 NbL07g14240.1 NbL07g15520.1 NbL07g16220.1 NbL07g17480.1 NbL08g01820.1 NbL08g02750.1 NbL08g02930.1 NbL08g03510.1 NbL08g03620.1 NbL08g03850.1 NbL08g03970.1 NbL08g04040.1 NbL08g06150.1 NbL08g06410.1 NbL08g06680.1 NbL08g06730.1 NbL08g07620.1 NbL08g08450.1 NbL08g08640.1 NbL08g09910.1 NbL08g10160.1 NbL08g11760.1 NbL08g12570.1 NbL08g13630.1 NbL08g13890.1 NbL08g15050.1 NbL08g15340.1 NbL08g18010.1 NbL08g18420.1 NbL08g19080.1 NbL08g19190.1 NbL09g00900.1 NbL09g02160.1 NbL09g02330.1 NbL09g02470.1 NbL09g04150.1 NbL09g05210.1 NbL09g07010.1 NbL09g09070.1 NbL09g10290.1 NbL09g10500.1 NbL09g11220.1 NbL09g13490.1 NbL09g15290.1 NbL09g15830.1 NbL09g17240.1 NbL09g19250.1 NbL09g19460.1 NbL09g20190.1 NbL09g21040.1 NbL09g21520.1 NbL09g23460.1 NbL10g00080.1 NbL10g03710.1 NbL10g04330.1 NbL10g04560.1 NbL10g05200.1 NbL10g06320.1 NbL10g07510.1 NbL10g07960.1 NbL10g08670.1 NbL10g08970.1 NbL10g11120.1 NbL10g11340.1 NbL10g11820.1 NbL10g13720.1 NbL10g14560.1 NbL10g14770.1 NbL10g16430.1 NbL10g18140.1 NbL10g18380.1 NbL10g19280.1 NbL10g19690.1 NbL10g21210.1 NbL10g22680.1 NbL10g23160.1 NbL10g23560.1 NbL10g24210.1 NbL11g00680.1 NbL11g00970.1 NbL11g01230.1 NbL11g01270.1 NbL11g01520.1 NbL11g01530.1 NbL11g02920.1 NbL11g03540.1 NbL11g03990.1 NbL11g05630.1 NbL11g08950.1 NbL11g08980.1 NbL11g09510.1 NbL11g10840.1 NbL11g11030.1 NbL11g11230.1 NbL11g12430.1 NbL11g13300.1 NbL11g15430.1 NbL11g16390.1 NbL11g16410.1 NbL11g17320.1 NbL11g18090.1 NbL11g21310.1 NbL11g21470.1 NbL11g21780.1 NbL11g21820.1 NbL11g22270.1 NbL11g22310.1 NbL11g23180.1 NbL11g24100.1 NbL12g00130.1 NbL12g01810.1 NbL12g02230.1 NbL12g02720.1 NbL12g02760.1 NbL12g04120.1 NbL12g04550.1 NbL12g06630.1 NbL12g07830.1 NbL12g09170.1 NbL12g10580.1 NbL12g12090.1 NbL12g12490.1 NbL12g12630.1 NbL12g12800.1 NbL12g13320.1 NbL12g13460.1 NbL12g14430.1 NbL12g14970.1 NbL12g15490.1 NbL12g17460.1 NbL12g18190.1 NbL12g18590.1 NbL12g19900.1 NbL12g20690.1 NbL12g22040.1 NbL12g22560.1 NbL13g00350.1 NbL13g01440.1 NbL13g02400.1 NbL13g03210.1 NbL13g03360.1 NbL13g04070.1 NbL13g05250.1 NbL13g08460.1 NbL13g09010.1 NbL13g09140.1 NbL13g10290.1 NbL13g11570.1 NbL13g13370.1 NbL13g14910.1 NbL13g18680.1 NbL13g19510.1 NbL13g23520.1 NbL13g24010.1 NbL13g24190.1 NbL13g24460.1 NbL13g26310.1 NbL13g26640.1 NbL13g26860.1 NbL13g27260.1 NbL13g27960.1 NbL14g02460.1 NbL14g02750.1 NbL14g08750.1 NbL14g08910.1 NbL14g09120.1 NbL14g09540.1 NbL14g09920.1 NbL14g11070.1 NbL14g11150.1 NbL14g12570.1 NbL14g14530.1 NbL14g14860.1 NbL14g15240.1 NbL14g15460.1 NbL14g16620.1 NbL14g16910.1 NbL14g17940.1 NbL14g21150.1 NbL14g21750.1 NbL14g21910.1 NbL15g00790.1 NbL15g01170.1 NbL15g02310.1 NbL15g04220.1 NbL15g05970.1 NbL15g06340.1 NbL15g06440.1 NbL15g07020.1 NbL15g07370.1 NbL15g07470.1 NbL15g09010.1 NbL15g13210.1 NbL15g14550.1 NbL15g14600.1 NbL15g17290.1 NbL15g18170.1 NbL15g19710.1 NbL15g21840.1 NbL15g21930.1 NbL15g23410.1 NbL15g23420.1 NbL15g23430.1 NbL15g25130.1 NbL15g25200.1 NbL16g01760.1 NbL16g02140.1 NbL16g04460.1 NbL16g05010.1 NbL16g05020.1 NbL16g06780.1 NbL16g07540.1 NbL16g07980.1 NbL16g09760.1 NbL16g10610.1 NbL16g12320.1 NbL16g13510.1 NbL16g14420.1 NbL16g15690.1 NbL16g17420.1 NbL16g17790.1 NbL16g17880.1 NbL16g18730.1 NbL16g18940.1 NbL16g19440.1 NbL16g20980.1 NbL16g23180.1 NbL16g23610.1 NbL16g23660.1 NbL16g23910.1 NbL16g24550.1 NbL16g24640.1 NbL16g25300.1 NbL16g25630.1 NbL16g26710.1 NbL17g01590.1 NbL17g02070.1 NbL17g02120.1 NbL17g02920.1 NbL17g03040.1 NbL17g03540.1 NbL17g03700.1 NbL17g03800.1 NbL17g05400.1 NbL17g07510.1 NbL17g08450.1 NbL17g08930.1 NbL17g10090.1 NbL17g14370.1 NbL17g14600.1 NbL17g15390.1 NbL17g15900.1 NbL17g16000.1 NbL17g16910.1 NbL17g17480.1 NbL17g18240.1 NbL17g20020.1 NbL17g20830.1 NbL17g21220.1 NbL17g21690.1 NbL17g25960.1 NbL18g00030.1 NbL18g00150.1 NbL18g00310.1 NbL18g00670.1 NbL18g00700.1 NbL18g01630.1 NbL18g02650.1 NbL18g04460.1 NbL18g05210.1 NbL18g05690.1 NbL18g07270.1 NbL18g07440.1 NbL18g07500.1 NbL18g09090.1 NbL18g09810.1 NbL18g09880.1 NbL18g10500.1 NbL18g10990.1 NbL18g12070.1 NbL18g13060.1 NbL18g17480.1 NbL19g00230.1 NbL19g04470.1 NbL19g04700.1 NbL19g07770.1 NbL19g07870.1 NbL19g09260.1 NbL19g09870.1 NbL19g13480.1 NbL19g14360.1 NbL19g14400.1 NbL19g14720.1 NbL19g17700.1 NbL19g18860.1 NbL19g22750.1 NbQ13g00370.1 NbQ14g10310.1 NbQ17g06050.1

This script appears to have some problems with the file:

import pandas as pd

orthofinder_output = "../OrthoFinder-res/Results_Jan23/Orthogroups/Orthogroups-fixed.txt"
orthogroups = pd.read_csv(orthofinder_output, sep=' ', header=None)

# Extract gene family information
expansions = {}
contractions = {}
for i, row in orthogroups.iterrows():
    if len(row) > 2:
        expansions[row[0]] = row[2:]
    else:
        contractions[row[0]] = row[1]

# Print expansions and contractions
print("Expansions:", expansions)
print("Contractions:", contractions)

I got the following error:

geneFamilyExpansionsContractions.py:
 20: DtypeWarning:
   Columns (2,3,4,5,6,...,964,965,966,967,968) have mixed types.
   Specify dtype option on import or set low_memory=False.
  orthogroups = pd.read_csv(orthofinder_output, sep=' ', header=None)
...
Expansions:
...
Name: 63241, Length: 967, dtype: object, 'OG0063242':
2      NaN
3      NaN
4      NaN
5      NaN
6      NaN
      ... 
964    NaN
965    NaN
966    NaN
967    NaN
968    NaN
Name: 63242, Length: 967, dtype: object}
Contractions: {}

How can it be fixed?

$\endgroup$

2 Answers 2

1
$\begingroup$

The answer is actually right there in the error message (warning message, actually):

DtypeWarning: Columns (2,... 968) have mixed types. Specify dtype option on import or set low_memory=False.

Setting low_memory-False makes the error go away:

orthogroups = pd.read_csv(orthofinder_output, sep=' ', header=None, low_memory=False)

According to this random blog post I found when I searched for "pandas DtypeWarning: have mixed types", this is because:

Pandas is really nice because instead of stopping altogether, it guesses which dtype a column has. However, by default, Pandas has the low_memory=True argument. This means that the CSV file gets split up in multiple chunks and the guess is done for every chunk, resulting in a column with multiple dtypes. Pandas is so kind to let you know it was confused and something might have happened.

The same blog also suggests that this isn't really the right approach, and that makes sense, since this is forcing Pandas to read the entire file and that might be a problem for huge files or low memory machines. So, reading on, the blog suggests the other option provided by the error, setting the dtype manually.

Now, I am afraid I have never used Pandas and don't really know much Python either, but a quick glance at the Pandas documentation suggests that the object data type is the most versatile and should work for everything, including strings. Since your file is nothing but a collection of strings, as far as I can tell, this seems like the best approach.

So I changed your script to this:

#!/usr/bin/env python3

import pandas as pd
import sys

orthofinder_output = sys.argv[1]
orthogroups = pd.read_csv(orthofinder_output, sep=' ', header=None, low_memory=False)
# Extract gene family information
expansions = {}
contractions = {}
for i, row in orthogroups.iterrows():
    if len(row) > 2:
        expansions[row[0]] = row[2:]
    else:
        contractions[row[0]] = row[1]

# Print expansions and contractions
print("Expansions:", expansions)
print("Contractions:", contractions)

And it seems to work just fine with no warnings shown.

$\endgroup$
0
$\begingroup$

The error right at the end is

Specify dtype option on import or set low_memory=False.

Edit Oh wait @terdon has already spotted this! :-D ... great minds and all that :)

In addition, the pandas print out it states ...

dtype object

This isn't good. Name: 63242, is a bit unusual BTW worth checking.

Looking at the values ...

NaN

... that is really not good.

What I think is happening is it's been assigned to a numerical datatype, but it's categorical data. Hence NaN because like the alphabet can be none-base 10 numbers, in data compression used a lot, but it's got to be consistent pattern (I genuinely don't know if pandas can spot/work in none-base 10 number systems). pandas is really C (probably C++) - its a static type backend - so those datatypes are the key to this.

It makes sense because your data is categorical data and NaN means its managed to confuse a large amount of it for e.g. int64.

Looking the documentation the alternative is

orthogroups = pd.read_csv(orthofinder_output, sep=' ', header=None, dtype='category')

The documentation is

low_memorybool, default True Internally process the file in chunks, resulting in lower memory use while parsing, but possibly mixed type inference. To ensure no mixed types either set False, or specify the type with the dtype parameter. Note that the entire file is read into a single DataFrame regardless, use the chunksize or iterator parameter to return the data in chunks. (Only valid with C parser).

What you want is to issue a df.dtypes command once it is loaded and the output will define the datatype per column,

'A'    int64
'B'    int32
'C'   category

If you have mixed dtypes it may (I stress may) be able to correct it via,

df.astype('int64') # this is not your data
df.astype('category') # this is your data

HOWEVER, I don't think this will work, because NaN indicates its trashed the data. Anyway the data is definitely categorical BTW.

Anyway, good question and interesting stuff.


As a bit of gossip, you can also switch to categorical ordered data I think this will work for dataframes, this example is Series. There's other ways to order the data BTW.

from pandas.api.types import CategoricalDtype
my_dtype = CategoricalDtype(\
    categories=[2, 1], ordered=True)
df.dtype(my_dtype)

The output is 1 and 2 now switch places because 2 < 1. Categorial datatypes look to be the right datatype for your inputs.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.